Abdelghani
Halimi

PhD Student in Computer Science, Data & AI

Télécom SudParis · Institut Polytechnique de Paris

Enhancing Clinical Decision-Making in Liver Transplantation Using Machine Learning

Portrait of Abdelghani Halimi
Télécom SudParis — Institut Polytechnique de Paris Chaire Innovation BOPA SAMOVAR Laboratory École Nationale Polytechnique d'Alger

Available for Post-Doc & Industry Roles — November 2026

About & Research

I am a PhD student at Télécom SudParis (SAMOVAR laboratory, ARMEDIA team), within Institut Polytechnique de Paris. My doctoral research focuses on developing machine learning-based clinical decision support tools for liver transplantation, leveraging large-scale longitudinal registry data derived from electronic health records.

My work spans the development of explainable risk scores for waitlist mortality prediction, recipient-donor matching optimization with emphasis on equity and fairness, and clinical validation in collaboration with healthcare professionals at AP-HP Paul-Brousse Hospital. I hold an engineering degree from the National Polytechnic School of Algiers and a Master 2 from IP Paris, where I ranked 1st in my class.

Machine Learning Explainable AI (XAI) Clinical Decision Support Survival Analysis Electronic Health Records Liver Transplantation Deep Learning Gait Analysis

News & Updates

  • May 2026 — My research on leveraging machine learning for liver transplant allocation was featured in an editorial piece by IMTech. [Read Article]
  • January 2026 — Participated in an immersion day for middle school students at Télécom SudParis. This event was organized as part of the DaTSHealth project, with the support of the ANR (Agence nationale de la recherche). [View on LinkedIn]
  • October 2024 — Participated in the "Fête de la Science" at Télécom SudParis, engaging with the public to share insights on our research and the impact of artificial intelligence in healthcare. [Read Article]
  • [June 2022] — Recognized as a laureate of the Excellence Scholarship awarded by the French Embassy in Algeria and the Institut Français d'Algérie (IFA), and invited to give a speech sharing my academic journey. [View Post]

Publications

8 peer-reviewed publications in journals, IEEE conferences, and international medical congresses

Journal Articles

[1]

Halimi, A. et al., 2026.

Explainable Mortality Prediction for Liver Transplant Candidates with Hepatocellular Carcinoma: A Supervised Clustering Approach.

Health Data Science (Science Partner Journal).

View Details
HCC study design and cohort flow Feature Rank Stability Heatmap Feature Effects on Model Predictions Mortality Risk Stratification Across Identified Patient Clusters

Explainable mortality prediction for HCC liver transplant candidates using supervised clustering. Identifies 7 risk clusters revealing different pathways to mortality — liver failure vs. aggressive tumor-driven profiles.

[2]

Halimi, A. et al., 2025.

Explainable machine learning for prognostic modeling of waitlist mortality in cirrhotic liver transplantation.

Computational and Structural Biotechnology Journal.

View Details
End-to-end ML workflow for non-HCC waitlist mortality Feature importance analysis ELTM vs. MELD: Mortality Risk in TX Patients Clinical Utility Comparison: ELTM vs. MELD Scores Global explainability summary

End-to-end explainable ML pipeline for prognostic modeling of waitlist mortality in cirrhotic liver transplantation, with global and local interpretability via SHAP-based analysis.

[3]

Halimi, A. et al., 2025.

A novel gait quality measure for characterizing pathological gait based on Hidden Markov Models.

Computers in Biology and Medicine, 184, p.109368.

View Details
Hidden Markov Model architecture for gait analysis HMM-based gait quality scores

Novel gait quality measure using Hidden Markov Models to characterize pathological gait deviations from statistically modeled normal gait patterns.

[4]

Hermez, L., Halimi, A. et al., 2023.

Clinical gait analysis: characterizing normal gait and pathological deviations due to neurological diseases.

Sensors, 23(14), p.6566.

View Details
Temporal Alignment of Normal Gait Cycles via DTW Normal vs pathological gait comparison Gait deviation quantification results

Clinical gait analysis using Dynamic Time Warping to characterize normal gait and quantify pathological deviations due to neurological diseases.

Conference Papers

[5]

Halimi, A. et al., 2025.

A Multi-Task Learning Framework For Mortality Prediction in Liver Transplant Candidates.

IEEE 38th International Symposium on Computer-Based Medical Systems (CBMS).

View Details
Multi-task learning architecture Feature effects on HCC branch predictions Feature effects on Non-HCC branch predictions MTL vs single-task performance comparison

Multi-task learning framework jointly predicting mortality for HCC and non-HCC liver transplant candidates, with shared representation learning specifically benefiting HCC.

[6]

Halimi, A. et al., 2025.

Development of an explainable machine learning model for predicting waitlist mortality in liver transplant candidates.

Journal of Hepatology 82 (2025): S385.

Rising Star Award — ILTS 2025 View Details
Feature importance ranking

Development of an explainable ML model for predicting 3-month waitlist mortality.

[7]

Halimi, A. et al., 2025.

Machine learning in predicting 3-month mortality for liver transplant candidates with HCC: a paradigm shift.

Journal of Hepatology 82 (2025): S384–S385.

View Details
Performance metrics for traditional scores and the ML models with different variable (vars) counts.

ML-based paradigm shift in predicting 3-month mortality for HCC liver transplant candidates, outperforming conventional scoring systems.

[8]

Halimi, A. et al., 2024.

Predicting Waitlist Mortality for Liver Transplant Candidates: A Comparative Analysis between Statistical Scores and Machine Learning Models.

E-Health and Bioengineering Conference (EHB), pp. 1–4. IEEE.

Honorable Mention View Details
Score performance comparison (3-Month mortality period)

Comparative analysis between traditional statistical scores (MELD, MELD-Na, MELD 3.0) and machine learning models for predicting waitlist mortality in liver transplant candidates.

Scientific Communications

Oral

European Society for Organ Transplantation (ESOT) Congress

June 2025 · London, UK

Poster

European Association for the Study of the Liver (EASL) Congress

May 2025 · Netherlands

Oral / Poster

International Liver Transplantation Society (ILTS) Congress

May 2025 · Singapore

Poster

IMT Colloquium, Télécom SudParis

May 2025

Poster

Engineering for Health (E4H) Annual Forum, IP Paris

November 2024

Poster

Fête de la Science "Océan de Savoirs", IP Paris

October 2024

Poster

Hi! Paris Summer School, HEC Paris

July 2024

Poster

PhD Students' Day, Université Paris-Saclay / IP Paris

June 2024

Peer Review Activities

I serve as a peer reviewer for international scientific journals, including:

Scientific Reports (Q1) BMC Medical Research Methodology (Q1)

Professional Experience

Oct 2023 — Oct 2026

PhD Thesis

SAMOVAR Lab, ARMEDIA Team · Télécom SudParis

Enhancing Clinical Decision-Making in Liver Transplantation Using Machine Learning

  • Development of ML-based risk scores using large-scale longitudinal clinical registry data from EHRs
  • Identification of key risk factors for waitlist mortality and post-transplant outcomes
  • Enhanced recipient-donor matching and organ allocation, prioritizing equity and fairness
  • Integration of explainability tools to build clinical trust
  • Validation using external retrospective data from French biomedical registries, in collaboration with healthcare professionals
View Research Highlights

Explainable ML for Non-HCC Waitlist Mortality (CSBJ 2025)

Non-HCC ML workflow Feature importance Global explainability

HCC Mortality Prediction via Supervised Clustering (HDS 2026)

HCC study design Feature effects on HCC model predictions Mortality risk stratification across identified patient clusters

Multi-Task Learning Framework (IEEE CBMS 2025)

MTL architecture MTL performance
Apr — Sep 2023

M2 Research Internship

SAMOVAR Lab, ARMEDIA Team · Télécom SudParis

Characterization of Normal and Pathological Gait Through Statistical and Deep Models

  • Derived Normal Gait Profiles using K-medoids with Dynamic Time Warping distance
  • Built a statistical model of normal gait using Hidden Markov Models
  • Learned normal gait representations through Autoencoder-based models
  • Proposed gait quality metrics quantifying pathological deviation from normal gait
View Research Highlights

Gait Analysis Overview

Motor impairment induced by neurological disorders Gait cycle representation

Dynamic Time Warping (DTW) Approach — Sensors 2023

Temporal alignment of normal gait cycles via DTW 3D DTW Metric for Normal and Pathological Gait Assessment DTW gait deviation

Hidden Markov Models (HMMs) — CBM 2025

HMM architecture HMM gait quality scores
Feb — Jul 2022

Engineering Research Internship

National Polytechnic School of Algiers (ENP)

Enhancing Deep Learning-based Classifiers using Out-of-Distribution Data Detection

  • Comparative evaluation of state-of-the-art OOD detection methods
  • Integration of OOD detection into deep learning classifiers on CIFAR and ImageNet
  • Proposed a novel OOD detection method outperforming baseline approaches
  • Developed a web-based interface for real-time testing and demonstration
View Research Highlights

OOD Detection for Deep Learning Classifiers

Training pipeline with in-distribution data OOD detection at test time

Proposed a novel Out-of-Distribution detection method integrated into deep learning classifiers, evaluated on CIFAR-10/100 and ImageNet benchmarks.

Education

Since Oct 2023

PhD in Computer Science, Data & AI

Institut Polytechnique de Paris

Sep 2022 — Sep 2023

M2 Information Processing & Data Exploitation (TRIED)

Institut Polytechnique de Paris

Ranked 1st / 22
Sep 2019 — Jul 2022

Engineering Degree & M2 in Electronics

National Polytechnic School of Algiers

Ranked 5th / 18
Sep 2017 — Sep 2019

Preparatory Classes in Science & Technology

National Polytechnic School of Algiers

Ranked 29th / 1,037
Jun 2017

Baccalaureate in Mathematics

Abdelhafid Boussouf High School, Algiers

Very Good — 16.93 / 20

Teaching & Supervision

70+ hours of teaching across 4 programs (in English and French). Designed and prepared lecture content, slides, laboratory sessions, and associated pedagogical materials (except for the Master 2 TRIED courses).

M2 BME

17h

École Polytechnique

  • Fundamentals of ML & Applications — 14h: PCA, t-SNE, UMAP, XAI (LIME, SHAP), clustering, ensemble methods
  • Data Preparation & Workflow — 3h: Imputation methods, outlier detection for biomedical data

M2 TRIED

23h

Télécom SudParis

  • Case Study 2 — 20h: End-to-end ML project on EEG-based classification of neurocognitive disorders
  • Pattern Recognition & Neural Methods — 3h: GMMs, entropy features, unsupervised clustering

M1 DATAPAC

18h

Télécom SudParis

  • AI for Data Science — 18h lectures: Supervised/unsupervised learning, deep learning, autoencoders, time-series modeling

Engineering Program - Year 2

13h

Télécom SudParis / ENSIIE

  • Digital Health / Pattern Recognition and Biometrics — 13h: Unsupervised learning, DTW clustering, gait analysis, ECG anomaly detection

Supervision

Engineering Year 2

180h

Télécom SudParis

  • Survival prediction in cirrhotic patients using tree-based ensemble models
  • Unsupervised learning for EEG-based patient profiling in neurocognitive disorders

M2 BME Student

Ongoing

École Polytechnique

  • Predicting post-graft survival using clinical registry data from EHRs

Awards & Honors

June 2025

BME Conference Fellowship

Engineering for Health (E4H), IP Paris

Competitive fellowship to attend and present at the 38th IEEE CBMS, funded by EUR-BERTIP (ANR 18EURE0002) under Plan France 2030.

May 2025

Rising Star Award

ILTS Congress 2025

Awarded for the abstract on explainable ML for predicting waitlist mortality, accepted for oral presentation at the ILTS Congress.

November 2024

Honorable Mention

IEEE EHB Young Researcher Contest

Research paper selected among the top 5 in the Young Researcher contest at the IEEE EHB Conference.

October 2023

PhD Funding — BOPA Innovation Chair

AP-HP, Institut Mines-Télécom, Univ. Paris-Saclay

Doctoral research funded by the BOPA Innovation Chair, a joint initiative of AP-HP, Institut Mines-Télécom and Université Paris-Saclay, based at Paul-Brousse Hospital. It brings together Inria, CNAM and leading med-tech partners.

January 2023

Merit Scholarship

SAMOVAR Laboratory

Scholarship awarded on merit and competition for a 6-month research internship at the SAMOVAR laboratory.

September 2022

Admission to PhD Track

Institut Polytechnique de Paris

Integrated M2–PhD program offering high-level research training to high-potential students aiming for international academic careers.

June 2022

IFA Excellence Scholarship

French Institute of Algeria / French Embassy

One of 20 scholarships awarded to welcome Algerian students in M2 programs at French higher education institutions, based on exceptional academic results.

Contact & Profiles

Email abdelghani.halimi@telecom-sudparis.eu
LinkedIn LinkedIn
Location

Évry-Courcouronnes, Essonne, France

Languages

French (C1) · English (C1) · Arabic (Native)

Google Scholar Google Scholar