Medical Data Science

Machine learning for critical care — predicting ICU feeding complications, phenotyping rare respiratory diseases, and enabling privacy-preserving medical data collaboration.

Headline result: Clinical prediction and privacy-preserving collaboration.

Papers: 17
Years: 2015–2026
Co-authors: 50

Overview

Critical care patients are among the most vulnerable — yet managing their nutrition, predicting complications, and coordinating care remains largely manual. We build machine learning systems that turn ICU data into actionable clinical decisions, improving patient outcomes through real-time prediction.

Research Areas

ICU Nutrition & Feeding Complications — Models trained on a decade of records from Rabin Medical Center that predict feeding intolerance and caloric deficit before symptoms appear, optimizing enteral nutrition per patient.
Rare Respiratory Disease Phenotyping — Unsupervised clustering and AI phenotypical analysis for NTM lung infections, accelerating diagnosis and predicting treatment response in pulmonology.
Privacy-Preserving Medical Data Collaboration — Federated learning and crowdsourcing frameworks enabling multi-hospital research without exposing patient records.

Technical Methods

Our pipeline combines gradient-boosted trees and deep learning for tabular clinical data, time-series models for continuous monitoring streams, and survival analysis for time-to-event outcomes. All models are validated prospectively against held-out hospital cohorts before clinical reporting.

Left: Real-time patient monitoring generating continuous data streams. Middle: Critical care enteral nutrition delivery. Right: AI-driven analysis of complex clinical data patterns.

People on this line

The 8 most frequent of 50 co-authors on this line since 2015.

Ranked by joint papers in this line. The bar under each name spans their years on it, 2015 to 2026. A ring marks a student of the lab (3 of 50 co-authors here). 42 more co-authors on the Lab page

Papers

17 publications, newest first.

All publications

2026

Machine Learning Tools for Predicting Pediatric Urinary Tract Infections Caused by ESBL-Producing Bacteria

Chen Hajaj, Shani Alkoby, Shai Ashkenazi, Avner Herman Cohen, Yael Reichenberg, Shiri Kushnir, and Vered Shkalim Zemer

JournalPediatric Infectious Disease Journal SJR Q1

DOI

Background: The prevalence of pediatric urinary tract infections (UTIs) caused by extended-spectrum β-lactamases (ESBL)-producing bacteria is increasing worldwide and is difficult to predict. As these infections require special antibiotic treatment, which is often not started empirically, they are associated with higher rates of intensive care unit admission, morbidity and prolonged hospitalization. We aimed to develop machine learning-based tools to aid pediatricians in predicting ESBL-positive UTIs and initiate appropriate empiric antibiotics. Methods: The electronic medical records of a large Health Maintenance Organization were searched for all children one month to 18 years of age with confirmed UTIs during January 1, 2010, to August 31, 2020. Data on demographics, clinical and laboratory information were retrieved, and following univariate analysis, machine learning-based tools were used to develop models to predict a UTI caused by an ESBL-producing bacterium. Results: A total of 35,830 pediatric UTI events comprised the study group. Age, sex, socioeconomic status, site of infection (community or hospital), prior antibiotic use, previous ESBL-positive UTI and the specific uropathogen were significantly associated with the rates of ESBL-positive infection. Using patients’ data available on presentation, the 5 models developed had a very high negative predictive value of 0.98, indicating strong rule-out performance for ESBL-positive UTIs. Conclusions: Our study indicates that machine learning models based on data available at UTI presentation may support clinicians in estimating the likelihood of ESBL-producing bacteria UTIs. Prospective studies are required to improve the models’ performance and determine their actual impact on clinical outcomes.
```
@article{shkalim2026esbl_uti_pidj,
  author = {Hajaj, Chen and Alkoby, Shani and Ashkenazi, Shai and Herman Cohen, Avner and Reichenberg, Yael and Kushnir, Shiri and Shkalim Zemer, Vered},
  title = {Machine Learning Tools for Predicting Pediatric Urinary Tract Infections Caused by ESBL-Producing Bacteria},
  journal = {Pediatric Infectious Disease Journal},
  year = {2026},
  doi = {10.1097/INF.0000000000005179}
}
```

2025

Survival Forest Models for ICU Mortality Prediction based on Nutrition and Clinical Factors

Anat Goldstein, Chen Hajaj, Pierre Singer, and Orit Raphaeli

JournalIEEE Access SJR Q1

DOI

Objective: This study investigates the predictive role of nutritional factors and gastrointestinal dysfunction in determining mortality among critically ill patients admitted to the intensive care unit (ICU). Our goal is to enhance mortality prediction by incorporating detailed nutritional and clinical data collected during the first 48 hours following ICU admission. Methods: The study included data from 2,942 patients admitted to Beilinson Hospital ICU. Variables collected comprised nutritional measures, gastrointestinal dysfunction symptoms, patient demographics, vital signs, respiratory parameters, and laboratory results. For each clinical variable, per-patient 48-hour summary statistics were derived. Three feature selection strategies were compared: stepwise Cox proportional hazards modeling, Random Forest feature importance, and Principal Component Analysis (PCA). For each feature set, we developed Random Survival Forest (RSF) models and evaluated predictive performance using six-fold cross-validation. Performance evaluation included the Concordance Index (CI), time-dependent AUC at 10 days, and Integrated Brier Score (IBS ≤10 days) for Random Survival Forest models, while Akaike Information Criterion (AIC) and CI were used to monitor model fit and discrimination during stepwise Cox proportional hazards feature selection. Results: The stepwise-selected feature set produced the most accurate model (CI=0.76) and achieved the strongest overall performance across complementary metrics (AUC10d=0.874, IBS ≤10 d=0.049). The Random Forest–selected feature set performed closely behind (CI=0.747; AUC10d=0.872; IBS ≤10 d=0.050). PCA-based feature reduction resulted in substantially poorer discrimination (CI=0.73) and higher prediction error (AUC10d=0.715; IBS ≤10 d=0.068). Across approaches, nutritional and gastrointestinal (GI) dysfunction variables, including nasogastric output, feeding tolerance, and timing of enteral nutrition, emerged as important predictors of early mortality. Conclusion: Our findings underscore the critical importance of early nutritional intervention and GI dysfunction management for ICU patient survival. The study demonstrates that targeted feature selection, which incorporates clinical insight, yields superior discriminative performance compared with purely data-driven reduction methods. Future research should validate these models externally and integrate real-time monitoring data to further refine clinical decision-making and improve patient outcomes in ICU settings.
```
@article{goldstein2025_survival_forest_icu,
  author = {Goldstein, Anat and Hajaj, Chen and Singer, Pierre and Raphaeli, Orit},
  title = {Survival Forest Models for ICU Mortality Prediction based on Nutrition and Clinical Factors},
  journal = {IEEE Access},
  year = {2025},
  doi = {10.1109/access.2025.3649629}
}
```

Predicting Onset and Progression of Neurodegenerative Diseases Using Blood Test Data and Machine Learning Models

Amir Glik, Chen Hajaj, Orit Raphaeli, and Anat Goldstein

PatentUS Patent Application Publication US 2025/0037877 A1 (Application No. 18/910,178)

Link

@misc{glik2025predicting,
  title = {Predicting Onset and Progression of Neurodegenerative Diseases Using Blood Test Data and Machine Learning Models},
  author = {Glik, Amir and Hajaj, Chen and Raphaeli, Orit and Goldstein, Anat},
  howpublished = {US Patent Application Publication US 2025/0037877 A1 (Application No. 18/910,178)},
  note = {Applicants: MOR Research Applications Ltd. and Ariel Scientific Innovations Ltd. Priority: US Provisional 63/456,555 (2023)},
  year = {2025},
  month = jan,
  url = {https://patents.justia.com/patent/20250037877},
  month_numeric = {1}
}

Selected: Routine Blood Count and Chemistry can predict AD dementia risk up to Ten years Ahead

Amir Glik, Omry Arbiv, Keshet Prado, Orit Raphaeli, Hayim Raclaw, Roy Kait, Ofri Kait, Anat Goldstein, and Chen Hajaj

AbstractConference abstract (AAIC 2025), Alzheimer’s & Dementia 21(S2):e102512

DOI

AD dementia risk prediction may help prevent 30% of cases by treating vascular risk factors years before clinical stage. Moreover, new disease modifying drugs are given only in the earliest clinical stages. Hence, there is a need for a tool that can flag high‐risk cognitive healthy (CH) subjects in order to improve early drug accessibility. The tool should be able to screen mass populations in a short period while maintaining low cost. The aim of this study was to develop such a tool. We interrogated a community cohort from the Clalit health care services. The cohort included 504,219 subjects, above the age of 45. Each subject had at least one routine blood count and basic chemistry and up to ten consecutive blood exams done on a yearly basis. The blood exams were done for various other reasons. After exclusion, there were 381,754 Cognitive Healthy (CH) subjects and 59,441 subjects diagnosed with AD (MCI or early dementia). AD Diagnosis was based on the 2011 NIA guidelines. The predictive ML tool was trained for different combinations of historical period (1 to 10y) and prediction horizons (1 to 10y). For each model, we calculated the accuracy, area under the curve (AUC), precision, recall, F1 score, false positive/negative rates, risk ratio (RR) and odds ratio (OR). We present herein some examples of the model results. For one year of blood exam history (aka one blood exam) and one year of horizon prediction the accuracy was 0.73, AUC 0.8, precision 0.28, recall 0.73, F1 score 0.4. For two years of blood exam history and three years of horizon prediction the accuracy was 0.73, AUC 0.84, precision 0.28, recall 0.82, F1 score 0.42. For two years of history and ten years of horizon the accuracy was 0.73, AUC 0.81, precision 0.19, recall 0.75, F1 score 0.3. Routine Blood count and chemistry, done for various other reasons, may enclose information concerning future risk for AD dementia. Using ML and AI sophisticated tools can facilitate the use of routine blood exams as a screening tool for AD dementia risk assessment.
```
@misc{glik2025routine,
  author = {Glik, Amir and Arbiv, Omry and Prado, Keshet and Raphaeli, Orit and Raclaw, Hayim and Kait, Roy and Kait, Ofri and Goldstein, Anat and Hajaj, Chen},
  title = {Routine Blood Count and Chemistry can predict AD dementia risk up to Ten years Ahead},
  howpublished = {Conference abstract (AAIC 2025), Alzheimer's \& Dementia 21(S2):e102512},
  year = {2025},
  doi = {10.1002/alz70856_102512},
  url = {https://doi.org/10.1002/alz70856_102512}
}
```

2024

Phenotypical Characteristics of Nontuberculous Mycobacterial Infection in Patients with Bronchiectasis

Assaf Frajman, Shimon Izhakian, Ori Mekiten, Ori Hadar, Ariel Lichtenstadt, Chen Hajaj, Shon Shchori, Moshe Heching, Dror Rosengarten, and Mordechai R Kramer

JournalRespiratory Research SJR Q1

DOI

Background The global mortality and morbidity rates of bronchiectasis patients due to nontuberculous mycobacteria (NTM) pulmonary infection are on a concerning upward trend. The aims of this study to identify the phenotype of NTM-positive individuals with bronchiectasis. Methods A retrospective single-center observational study was conducted in adult patients with bronchiectasis who underwent bronchoscopy in 2007-2020. Clinical, laboratory, pulmonary function, and radiological data were compared between patients with a positive or negative NTM culture. Results Compared to the NTM-negative group (n=677), the NTM-positive group (n=94) was characterized (P ≤0.05 for all) by older age, greater proportion of females, and higher rates of gastroesophageal reflux disease and muco-active medication use; lower body mass index, serum albumin level, and lymphocyte and eosinophil counts; lower values of forced expiratory volume in one second, forced vital capacity, and their ratio, and lower diffusing lung capacity for carbon monoxide; higher rates of bronchiectasis in both lungs and upper lobes and higher number of involved lobes; and more exacerbations in the year prior bronchoscopy. On multivariate analysis, older age (OR 1.05, 95% CI 1.02-1.07, P=0.001), lower body mass index (OR 1.16, 95% CI 1.16-1.07, P <0.001), and increased number of involved lobes (OR 1.26, 95% CI 1.01-1.44, P=0.04) were associated with NTM infection. Conclusions Patients with bronchiectasis and NTM pulmonary infection are more likely to be older and female with more severe clinical, laboratory, pulmonary function, and radiological parameters than those without NTM infection. This phenotype can be used for screening patients with suspected NTM disease.
```
@article{frajman2024phenotypical,
  title = {Phenotypical Characteristics of Nontuberculous Mycobacterial Infection in Patients with Bronchiectasis},
  author = {Frajman, Assaf and Izhakian, Shimon and Mekiten, Ori and Hadar, Ori and Lichtenstadt, Ariel and Hajaj, Chen and Shchori, Shon and Heching, Moshe and Rosengarten, Dror and Kramer, Mordechai R},
  journal = {Respiratory Research},
  volume = {25},
  number = {1},
  pages = {278},
  year = {2024},
  publisher = {Springer},
  doi = {10.1186/s12931-024-02904-0}
}
```

2023

Protein Intake and Clinical Outcomes of Enterally Fed Critically Ill Patients

Orit Raphaeli, Liran Statlender, C Hajaj, Itai Ben David, and Pierre Singer

ConferenceConference abstract, Clinical Nutrition ESPEN 54:534

DOI

@misc{raphaeli2023protein,
  journal = {Clinical Nutrition ESPEN},
  howpublished = {Conference abstract, Clinical Nutrition ESPEN 54:534},
  title = {Protein Intake and Clinical Outcomes of Enterally Fed Critically Ill Patients},
  author = {Raphaeli, Orit and Statlender, Liran and Hajaj, C and Ben David, Itai and Singer, Pierre},
  volume = {54},
  pages = {534},
  year = {2023},
  publisher = {Elsevier},
  doi = {10.1016/j.clnesp.2022.09.224}
}

Using Machine-Learning to Assess the Prognostic Value of Early Enteral Feeding Intolerance in Critically Ill Patients: A Retrospective Study

Orit Raphaeli, Liran Statlender, Chen Hajaj, Itai Ben David, Anat Goldstein, Eyal Robinson, and Pierre Singer

JournalNutrients SJR Q1

DOI

Background: The association between gastrointestinal intolerance during early enteral nutrition (EN) and adverse clinical outcomes in critically ill patients is controversial. We aimed to assess the prognostic value of enteral feeding intolerance (EFI) markers during early ICU stays and to predict early EN failure using a machine learning (ML) approach. Methods: We performed a retrospective analysis of data from adult patients admitted to Beilinson Hospital ICU between January 2011 and December 2018 for more than 48 h and received EN. Clinical data, including demographics, severity scores, EFI markers, and medications, along with 72 h after admission, were analyzed by ML algorithms. Prediction performance was assessed by the area under the receiver operating characteristics (AUCROC) of a ten-fold cross-validation set. Results: The datasets comprised 1584 patients. The means of the cross-validation AUCROCs for 90-day mortality and early EN failure were 0.73 (95% CI 0.71–0.75) and 0.71 (95% CI 0.67–0.74), respectively. Gastric residual volume above 250 mL on the second day was an important component of both prediction models. Conclusions: ML underlined the EFI markers that predict poor 90-day outcomes and early EN failure and supports early recognition of at-risk patients. Results have to be confirmed in further prospective and external validation studies.
```
@article{raphaeli2023using,
  title = {Using Machine-Learning to Assess the Prognostic Value of Early Enteral Feeding Intolerance in Critically Ill Patients: A Retrospective Study},
  author = {Raphaeli, Orit and Statlender, Liran and Hajaj, Chen and Ben David, Itai and Goldstein, Anat and Robinson, Eyal and Singer, Pierre},
  journal = {Nutrients},
  volume = {15},
  number = {12},
  pages = {2705},
  year = {2023},
  publisher = {MDPI},
  doi = {10.3390/nu15122705}
}
```

2022

Using the Cardio-Vascular Index (CVRI) to Predict Mortality in Septic Shock

Orit Raphaeli, Itai Ben David, C Hajaj, Liran Statlender, Anat Goldstein, Elad Chen, Pierre Singer, and U Gabbay

ConferenceConference abstract, ISICEM 2022

@misc{raphaeli2022using,
  title = {Using the Cardio-Vascular Index (CVRI) to Predict Mortality in Septic Shock},
  howpublished = {Conference abstract, ISICEM 2022},
  author = {Raphaeli, Orit and Ben David, Itai and Hajaj, C and Statlender, Liran and Goldstein, Anat and Chen, Elad and Singer, Pierre and Gabbay, U},
  year = {2022}
}

A Method for Evaluating the Performance of a Novel Percutaneous Left Ventricular Assist Device and Optimizing Its Design

Bar Vinder, Chen Hajaj, and Idit Avrahami

WorkshopVirtual Physiological Human Conference (VPH 2022) Student-led

@inproceedings{vinder2022method,
  title = {A Method for Evaluating the Performance of a Novel Percutaneous Left Ventricular Assist Device and Optimizing Its Design},
  author = {Vinder, Bar and Hajaj, Chen and Avrahami, Idit},
  booktitle = {Virtual Physiological Human Conference (VPH 2022)},
  year = {2022}
}

2021

Prediction Model for the Spread of the COVID-19 Outbreak in the Global Environment

Ron S. Hirschprung and Chen Hajaj

JournalHeliyon SJR Q1

DOI

COVID-19 has long become a worldwide pandemic. It is responsible for the death of over two million people and posed an economic recession. This paper studies the spread pattern of COVID-19, aiming to establish a prediction model for this event. We harness Data Mining and Machine Learning methodologies to train regression models to predict the number of confirmed cases in a spatial-temporal space. We introduce an innovative concept ‒ the Center of Infection Mass (CoIM) ‒ adapted from the field of physics. We empirically evaluated our model on western European countries, based on the CoIM index and other features, and showed that a relatively high accurate prediction of the spread can be obtained. Our contribution is twofold: first, we introduced a prediction methodology and proved empirically that a prediction can be made even to the range of over a month; second, we showed promise in adopting the CoIM index to prediction models, when models that adopt the CoIM yield significantly better results than those that discard it. By applying our model, and better controlling the inherent tradeoff between life-saving and economy, we believe that decision-makers can take close to optimal measures. Thus, this methodology may contribute to public welfare.
```
@article{hirschprung2021prediction,
  title = {Prediction Model for the Spread of the COVID-19 Outbreak in the Global Environment},
  author = {Hirschprung, Ron S. and Hajaj, Chen},
  journal = {Heliyon},
  volume = {7},
  number = {7},
  year = {2021},
  publisher = {Elsevier},
  doi = {10.1016/j.heliyon.2021.e07416}
}
```

Feeding Intolerance as a Predictor of Clinical Outcomes in Critically Ill Patients: A Machine Learning Approach

Orit Raphaeli, C Hajaj, Itai Ben David, Anat Goldstein, Elad Chen, and Pierre Singer

ConferenceConference abstract, Clinical Nutrition ESPEN 46:S546–S547

DOI

@misc{raphaeli2021feeding,
  journal = {Clinical Nutrition ESPEN},
  howpublished = {Conference abstract, Clinical Nutrition ESPEN 46:S546--S547},
  title = {Feeding Intolerance as a Predictor of Clinical Outcomes in Critically Ill Patients: A Machine Learning Approach},
  author = {Raphaeli, Orit and Hajaj, C and Ben David, Itai and Goldstein, Anat and Chen, Elad and Singer, Pierre},
  volume = {46},
  pages = {S546--S547},
  year = {2021},
  publisher = {Elsevier},
  doi = {10.1016/j.clnesp.2021.09.020}
}

Using machine learning to support early prediction of feeding intolerance in critically ill patients

Orit Raphaeli, C Hajaj, Itai Ben David, Anat Goldstein, Elad Chen, and Pierre Singer

ConferenceConference abstract, ESICM LIVES 2021

BACKGROUND: The association between gastrointestinal intolerance during early enteral nutrition (EN) and adverse clinical outcomes in critically ill patients is controversial. We aimed to assess the prognostic value of enteral feeding intolerance (EFI) markers during early ICU stays and to predict early EN failure using a machine learning (ML) approach. METHODS: We performed a retrospective analysis of data from adult patients admitted to Beilinson Hospital ICU between January 2011 and December 2018 for more than 48 h and received EN. Clinical data, including demographics, severity scores, EFI markers, and medications, along with 72 h after admission, were analyzed by ML algorithms. Prediction performance was assessed by the area under the receiver operating characteristics (AUCROC) of a ten-fold cross-validation set. RESULTS: The datasets comprised 1584 patients. The means of the cross-validation AUCROCs for 90-day mortality and early EN failure were 0.73 (95% CI 0.71-0.75) and 0.71 (95% CI 0.67-0.74), respectively. Gastric residual volume above 250 mL on the second day was an important component of both prediction models. CONCLUSIONS: ML underlined the EFI markers that predict poor 90-day outcomes and early EN failure and supports early recognition of at-risk patients. Results have to be confirmed in further prospective and external validation studies.
```
@misc{raphaeli2021using,
  title = {Using machine learning to support early prediction of feeding intolerance in critically ill patients},
  howpublished = {Conference abstract, ESICM LIVES 2021},
  author = {Raphaeli, Orit and Hajaj, C and Ben David, Itai and Goldstein, Anat and Chen, Elad and Singer, Pierre},
  year = {2021}
}
```

Using Machine Learning to Compare Gastric Residual Volume Thresholds as Predictors of Clinical Outcomes in Critically Ill Patients

Orit Raphaeli, Chen Hajaj, Itai Ben David, Anat Goldstein, Elad Chen, and Pierre Singer

ConferenceInternational Symposium on Intensive Care and Emergency Medicine (ISICEM)

@inproceedings{raphaeli2021gastric,
  title = {Using Machine Learning to Compare Gastric Residual Volume Thresholds as Predictors of Clinical Outcomes in Critically Ill Patients},
  author = {Raphaeli, Orit and Hajaj, Chen and Ben David, Itai and Goldstein, Anat and Chen, Elad and Singer, Pierre},
  booktitle = {International Symposium on Intensive Care and Emergency Medicine (ISICEM)},
  year = {2021}
}

2018

A crowdsourcing framework for medical data sets

Cheng Ye, Joseph R. Coco, Anna Epishova, Chen Hajaj, Henry Bogardus, Laurie Novak, Joshua Denny, Yevgeniy Vorobeychik, Thomas Lasko, Bradley Malin, and others

ConferenceAMIA Joint Summits on Translational Science Proceedings

Crowdsourcing services like Amazon Mechanical Turk allow researchers to ask questions to crowds of workers and quickly receive high quality labeled responses. However, crowds drawn from the general public are not suitable for labeling sensitive and complex data sets, such as medical records, due to various concerns. Major challenges in building and deploying a crowdsourcing system for medical data include, but are not limited to: managing access rights to sensitive data and ensuring data privacy controls are enforced; identifying workers with the necessary expertise to analyze complex information; and efficiently retrieving relevant information in massive data sets. In this paper, we introduce a crowdsourcing framework to support the annotation of medical data sets. We further demonstrate a workflow for crowdsourcing clinical chart reviews including (1) the design and decomposition of research questions; (2) the architecture for storing and displaying sensitive data; and (3) the development of tools to support crowd workers in quickly analyzing information from complex data sets.
```
@inproceedings{ye2018crowdsourcing,
  title = {A crowdsourcing framework for medical data sets},
  booktitle = {AMIA Joint Summits on Translational Science Proceedings},
  author = {Ye, Cheng and Coco, Joseph R. and Epishova, Anna and Hajaj, Chen and Bogardus, Henry and Novak, Laurie and Denny, Joshua and Vorobeychik, Yevgeniy and Lasko, Thomas and Malin, Bradley and others},
  volume = {2018},
  pages = {273},
  year = {2018}
}
```

Selected: Crowdsourcing Clinical Chart Reviews

Joseph R. Coco, Cheng Ye, Chen Hajaj, Yevgeniy Vorobeychik, Joshua Denny, Laurie Novak, Bradley Malin, Thomas Lasko, and D. Fabbri

ConferenceAMIA Annual Symposium Proceedings

@inproceedings{coco2018crowdsourcing,
  author = {Coco, Joseph R. and Ye, Cheng and Hajaj, Chen and Vorobeychik, Yevgeniy and Denny, Joshua and Novak, Laurie and Malin, Bradley and Lasko, Thomas and Fabbri, D.},
  title = {Crowdsourcing Clinical Chart Reviews},
  booktitle = {AMIA Annual Symposium Proceedings},
  year = {2018}
}

2015

Automatic Anatomical Shape Correspondence and Alignment Using Mesh Features

Tal Darom, Yaniv Gur, Chen Hajaj, and Yosi Keller

ConferenceIEEE International Symposium on Biomedical Imaging (ISBI)

DOI

In this work, we propose a fully automatic and computationally efficient group registration approach for sets of three-dimensional models represented as mesh objects. Our approach is based on agglomerating the set of pairwise model-to-model rigid registrations by a robust spectral synchronization scheme. The pairwise registration is computed using spectral graph matching applied to meshes via the LD-SIFT local mesh features. We applied the proposed scheme to sets of subcortical surfaces, and it was shown to provide accurate and robust registration results.
```
@inproceedings{darom2015automatic,
  title = {Automatic Anatomical Shape Correspondence and Alignment Using Mesh Features},
  author = {Darom, Tal and Gur, Yaniv and Hajaj, Chen and Keller, Yosi},
  booktitle = {IEEE International Symposium on Biomedical Imaging (ISBI)},
  pages = {1530--1534},
  year = {2015},
  doi = {10.1109/isbi.2015.7164169}
}
```
Strategy-Proof and Efficient Kidney Exchange Using a Credit Mechanism

Chen Hajaj, John P Dickerson, Avinatan Hassidim, Tuomas Sandholm, and David Sarne

ConferenceAAAI Conference on Artificial Intelligence (AAAI) CORE A*

DOI

We present a credit-based matching mechanism for dynamic barter markets—and kidney exchange in particular—that is both strategy proof and efficient, that is, it guarantees truthful disclosure of donor-patient pairs from the transplant centers and results in the maximum global matching. Furthermore, the mechanism is individually rational in the sense that, in the long run, it guarantees each transplant center more matches than the center could have achieved alone. The mechanism does not require assumptions about the underlying distribution of compatibility graphs—a nuance that has previously produced conflicting results in other aspects of theoretical kidney exchange. Our results apply not only to matching via 2-cycles: the matchings can also include cycles of any length and altruist-initiated chains, which is important at least in kidney exchanges. The mechanism can also be adjusted to guarantee immediate individual rationality at the expense of economic efficiency, while preserving strategy proofness via the credits. This circumvents a well-known impossibility result in static kidney exchange concerning the existence of an individually rational, strategy-proof, and maximal mechanism. We show empirically that the mechanism results in significant gains on data from a national kidney exchange that includes 59% of all US transplant centers.
```
@inproceedings{hajaj2015strategy,
  doi = {10.1609/aaai.v29i1.9322},
  title = {Strategy-Proof and Efficient Kidney Exchange Using a Credit Mechanism},
  author = {Hajaj, Chen and Dickerson, John P and Hassidim, Avinatan and Sandholm, Tuomas and Sarne, David},
  booktitle = {AAAI Conference on Artificial Intelligence (AAAI)},
  year = {2015}
}
```

Work on this

The lab takes on PhD and MSc students and works with academic and industry partners on this line of research. Write with a short note on your background and what you would like to work on.

Email Chen about this project Code and datasets