Abdur Rasool
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Abdur Rasool has a total of 99 co-authors. Here are the top co-authors by number of shared publications:
- Qingshan Jiang
Shenzhen Key Laboratory for High Performance Data Mining, Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences, Shenzhen, China
12 shared publications - Saba Aslam
Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen,518055, China, China
10 shared publications - Hui Chen
School of Artificial Intelligence, Shenzhen Polytechnic University, Shenzhen, China
8 shared publications - Qiang Qu
Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen 518055, China
6 shared publications - Dennis Paul Wall
Department of Pediatrics, Division of Clinical Informatics, Stanford University School of Medicine, Palo Alto, CA 94305, USA
4 shared publications - Marjan Kamyab
School of Computer Science and Technology, Donghua University, Shanghai, China
4 shared publications
Latest Publications
- Open Access
- 4 August 2026
- Published by MDPI AG in Preprints.org
GCE: A Framework for Interpretable Nonlinear Hazard Modeling in Cardiac Sarcoma Survival Using SEER Data
Cardiac sarcoma is a rare aggressive malignancy for which survival prediction is limited by small cohorts, censoring, nonlinear prognostic effects, and incomplete interpretability. We developed the GRU–CoxPH Ensemble (GCE), a weighted late-fusion survival framework combining a single-step GRU static-covariate learner with Cox proportional hazards (CoxPHs). Using SEER Research Plus data, 727 eligible patients were identified from 41 source variables, and 27 raw SEER predictors were retained after leakage exclusion and Cox-LASSO/Random Survival Forest screening. Outcome, follow-up, cause-of-death, identifier, and endpoint-derived fields were removed before preprocessing; imputation, encoding, scaling, feature selection, tuning, and ensemble-weight selection were performed within training folds only. In leakage-free 10-fold out-of-fold evaluation, the GCE achieved high cohort-specific internal discrimination that is likely optimistic and requires external validation (mean C-index =0.9198; IBS = 0.03641), whereas CoxPH showed lower discrimination (C-index = 0.8605) but better IBS calibration (0.03378). Thus, the GCE is better for discrimination, under internal validation, and is not calibration-superior. SHAP provided post hoc descriptive association summaries for the final ensemble risk score. These internal results support the GCE only as a research framework for cardiac sarcoma survival-risk modeling; external validation and recalibration are required before any patient-level clinical consideration.
- 4 August 2026
- Published by MDPI AG in Preprints.org
Open AccessGCE: A Framework for Interpretable Nonlinear Hazard Modeling in Cardiac Sarcoma Survival Using SEER Data
Cardiac sarcoma is a rare aggressive malignancy for which survival prediction is limited by small cohorts, censoring, nonlinear prognostic effects, and incomplete interpretability. We developed the GRU–CoxPH Ensemble (GCE), a weighted late-fusion survival framework combining a single-step GRU static-covariate learner with Cox proportional hazards (CoxPHs). Using SEER Research Plus data, 727 eligible patients were identified from 41 source variables, and 27 raw SEER predictors were retained after leakage exclusion and Cox-LASSO/Random Survival Forest screening. Outcome, follow-up, cause-of-death, identifier, and endpoint-derived fields were removed before preprocessing; imputation, encoding, scaling, feature selection, tuning, and ensemble-weight selection were performed within training folds only. In leakage-free 10-fold out-of-fold evaluation, the GCE achieved high cohort-specific internal discrimination that is likely optimistic and requires external validation (mean C-index =0.9198; IBS = 0.03641), whereas CoxPH showed lower discrimination (C-index = 0.8605) but better IBS calibration (0.03378). Thus, the GCE is better for discrimination, under internal validation, and is not calibration-superior. SHAP provided post hoc descriptive association summaries for the final ensemble risk score. These internal results support the GCE only as a research framework for cardiac sarcoma survival-risk modeling; external validation and recalibration are required before any patient-level clinical consideration.
- Open Access
- 1 August 2026
- Published by MDPI AG in Bioengineering
- Vol. 13 (8) , 891
GCE: A Framework for Interpretable Nonlinear Hazard Modeling in Cardiac Sarcoma Survival Using SEER Data
Cardiac sarcoma is a rare aggressive malignancy for which survival prediction is limited by small cohorts, censoring, nonlinear prognostic effects, and incomplete interpretability. We developed the GRU–CoxPH Ensemble (GCE), a weighted late-fusion survival framework combining a single-step GRU static-covariate learner with Cox proportional hazards (CoxPHs). Using SEER Research Plus data, 727 eligible patients were identified from 41 source variables, and 27 raw SEER predictors were retained after leakage exclusion and Cox-LASSO/Random Survival Forest screening. Outcome, follow-up, cause-of-death, identifier, and endpoint-derived fields were removed before preprocessing; imputation, encoding, scaling, feature selection, tuning, and ensemble-weight selection were performed within training folds only. In leakage-free 10-fold out-of-fold evaluation, the GCE achieved high cohort-specific internal discrimination that is likely optimistic and requires external validation (mean C-index =0.9198; IBS = 0.03641), whereas CoxPH showed lower discrimination (C-index = 0.8605) but better IBS calibration (0.03378). Thus, the GCE is better for discrimination, under internal validation, and is not calibration-superior. SHAP provided post hoc descriptive association summaries for the final ensemble risk score. These internal results support the GCE only as a research framework for cardiac sarcoma survival-risk modeling; external validation and recalibration are required before any patient-level clinical consideration.
- 1 August 2026
- Published by MDPI AG in Bioengineering
- Vol. 13 (8) , 891
Open AccessGCE: A Framework for Interpretable Nonlinear Hazard Modeling in Cardiac Sarcoma Survival Using SEER Data
Cardiac sarcoma is a rare aggressive malignancy for which survival prediction is limited by small cohorts, censoring, nonlinear prognostic effects, and incomplete interpretability. We developed the GRU–CoxPH Ensemble (GCE), a weighted late-fusion survival framework combining a single-step GRU static-covariate learner with Cox proportional hazards (CoxPHs). Using SEER Research Plus data, 727 eligible patients were identified from 41 source variables, and 27 raw SEER predictors were retained after leakage exclusion and Cox-LASSO/Random Survival Forest screening. Outcome, follow-up, cause-of-death, identifier, and endpoint-derived fields were removed before preprocessing; imputation, encoding, scaling, feature selection, tuning, and ensemble-weight selection were performed within training folds only. In leakage-free 10-fold out-of-fold evaluation, the GCE achieved high cohort-specific internal discrimination that is likely optimistic and requires external validation (mean C-index =0.9198; IBS = 0.03641), whereas CoxPH showed lower discrimination (C-index = 0.8605) but better IBS calibration (0.03378). Thus, the GCE is better for discrimination, under internal validation, and is not calibration-superior. SHAP provided post hoc descriptive association summaries for the final ensemble risk score. These internal results support the GCE only as a research framework for cardiac sarcoma survival-risk modeling; external validation and recalibration are required before any patient-level clinical consideration.
- 1 August 2026
- Published by Elsevier in Applied Soft Computing
- Vol. 203
Spatio-temporal feature fusion with cross-frequency coupling for interpretable Parkinson’s detection using EEG
- 1 August 2026
- Published by Elsevier in Applied Soft Computing
- Vol. 203
Spatio-temporal feature fusion with cross-frequency coupling for interpretable Parkinson’s detection using EEG
- 28 July 2026
- Published by Springer Nature
- p. 3-16
Introduction to DNA Storage Technology
- Top Cited Papers
- 4 April 2026
- Published by Elsevier in International Journal of Medical Informatics
- Vol. 214, 106417
Challenges in translating AI-driven ASD/ADHD diagnosis: A methodological systematic review
- 4 April 2026
- Published by Elsevier in International Journal of Medical Informatics
- Vol. 214, 106417
Top Cited PapersChallenges in translating AI-driven ASD/ADHD diagnosis: A methodological systematic review
- Top Cited Papers
- 3 March 2026
- Published by Elsevier in Computers and Electrical Engineering
- Vol. 133
Attention-guided electroencephalography fusion with fractal and nonlinear dynamics decodes early Alzheimer’s biomarkers
- 3 March 2026
- Published by Elsevier in Computers and Electrical Engineering
- Vol. 133
Top Cited PapersAttention-guided electroencephalography fusion with fractal and nonlinear dynamics decodes early Alzheimer’s biomarkers
- Open Access
- 21 October 2025
- Published by JMIR Publications Inc. in JMIR AI
- Vol. 4, e75030
Aiding Large Language Models Using Clinical Scoresheets for Neurobehavioral Diagnostic Classification From Text: Algorithm Development and Validation
Background: Large language models (LLMs) have demonstrated the ability to perform complex tasks traditionally requiring human intelligence. However, their use in automated diagnostics for psychiatry and behavioral sciences remains under-studied. Objective: This study aimed to evaluate whether incorporating structured clinical assessment scales improved the diagnostic performance of LLM-based chatbots for neuropsychiatric conditions (we evaluated autism spectrum disorder, aphasia, and depression datasets) across two prompting strategies: (1) direct diagnosis and (2) code generation. We aimed to contextualize LLM-based diagnostic performance by benchmarking it against prior work that applied traditional machine learning classifiers to the same datasets, allowing us to assess whether LLMs offer competitive or complementary capabilities in clinical classification tasks. Methods: We tested two approaches using ChatGPT, Gemini, and Claude models: (1) direct diagnostic querying and (2) execution of chatbot-generated code for classification. Three diagnostic datasets were used: ASDBank (autism spectrum disorder), AphasiaBank (aphasia), and Distress Analysis Interview Corpus-Wizard-of-Oz interviews (depression and related conditions). Each approach was evaluated with and without the aid of clinical assessment scales. Performance was compared to existing machine learning benchmarks on these datasets. Results: Across all 3 datasets, incorporating clinical assessment scales led to little improvement in performance, and results remained inconsistent and generally below those reported in previous studies. On the AphasiaBank dataset, the direct diagnosis approach using ChatGPT with GPT-4 produced a low F1-score of 65.6% and specificity of 33%. The code generation method improved results, with ChatGPT with GPT-4o reaching an F1-score of 81.4%, specificity of 78.6%, and sensitivity of 84.3%. ChatGPT with GPT-o3 and Gemini 2.5 Pro performed even better, with F1-scores of 86.5% and 84.3%, respectively. For the ASDBank dataset, direct diagnosis results were lower, with F1-scores of 56% for ChatGPT with GPT-4 and 54% for ChatGPT with GPT-4o. Under code generation, ChatGPT with GPT-o3 reached 67.9%, and Claude 3.5 performed reasonably well with 60%. Gemini 2.5 Pro failed to respond under this assessment condition. In the Distress Analysis Interview Corpus-Wizard-of-Oz dataset, direct diagnosis yielded high accuracy (70.9%) but poor F1-scores of 8% using ChatGPT with GPT-4o. Code generation improved specificity—88.6% with ChatGPT with GPT-4o—but F1-scores remained low overall. These findings suggest that, while clinical scales may help structure outputs, prompting alone remains insufficient for consistent diagnostic accuracy. Conclusions: Current LLM-based chatbots, when prompted naively, underperform on psychiatric and behavioral diagnostic tasks compared to specialized machine learning models. Clinical assessment scales might modestly aid chatbot performance, but more sophisticated prompt engineering and domain integration are likely required to reach clinically actionable standards.
- 21 October 2025
- Published by JMIR Publications Inc. in JMIR AI
- Vol. 4, e75030
Open AccessAiding Large Language Models Using Clinical Scoresheets for Neurobehavioral Diagnostic Classification From Text: Algorithm Development and Validation
Background: Large language models (LLMs) have demonstrated the ability to perform complex tasks traditionally requiring human intelligence. However, their use in automated diagnostics for psychiatry and behavioral sciences remains under-studied. Objective: This study aimed to evaluate whether incorporating structured clinical assessment scales improved the diagnostic performance of LLM-based chatbots for neuropsychiatric conditions (we evaluated autism spectrum disorder, aphasia, and depression datasets) across two prompting strategies: (1) direct diagnosis and (2) code generation. We aimed to contextualize LLM-based diagnostic performance by benchmarking it against prior work that applied traditional machine learning classifiers to the same datasets, allowing us to assess whether LLMs offer competitive or complementary capabilities in clinical classification tasks. Methods: We tested two approaches using ChatGPT, Gemini, and Claude models: (1) direct diagnostic querying and (2) execution of chatbot-generated code for classification. Three diagnostic datasets were used: ASDBank (autism spectrum disorder), AphasiaBank (aphasia), and Distress Analysis Interview Corpus-Wizard-of-Oz interviews (depression and related conditions). Each approach was evaluated with and without the aid of clinical assessment scales. Performance was compared to existing machine learning benchmarks on these datasets. Results: Across all 3 datasets, incorporating clinical assessment scales led to little improvement in performance, and results remained inconsistent and generally below those reported in previous studies. On the AphasiaBank dataset, the direct diagnosis approach using ChatGPT with GPT-4 produced a low F1-score of 65.6% and specificity of 33%. The code generation method improved results, with ChatGPT with GPT-4o reaching an F1-score of 81.4%, specificity of 78.6%, and sensitivity of 84.3%. ChatGPT with GPT-o3 and Gemini 2.5 Pro performed even better, with F1-scores of 86.5% and 84.3%, respectively. For the ASDBank dataset, direct diagnosis results were lower, with F1-scores of 56% for ChatGPT with GPT-4 and 54% for ChatGPT with GPT-4o. Under code generation, ChatGPT with GPT-o3 reached 67.9%, and Claude 3.5 performed reasonably well with 60%. Gemini 2.5 Pro failed to respond under this assessment condition. In the Distress Analysis Interview Corpus-Wizard-of-Oz dataset, direct diagnosis yielded high accuracy (70.9%) but poor F1-scores of 8% using ChatGPT with GPT-4o. Code generation improved specificity—88.6% with ChatGPT with GPT-4o—but F1-scores remained low overall. These findings suggest that, while clinical scales may help structure outputs, prompting alone remains insufficient for consistent diagnostic accuracy. Conclusions: Current LLM-based chatbots, when prompted naively, underperform on psychiatric and behavioral diagnostic tasks compared to specialized machine learning models. Clinical assessment scales might modestly aid chatbot performance, but more sophisticated prompt engineering and domain integration are likely required to reach clinically actionable standards.
- Open Access
- 30 June 2025
- Published by Institute of Central Computation and Knowledge Inc. in Journal of Artificial Intelligence in Bioinformatics
- Vol. 1 (1) , 41-50
RFS-codec: A Novel Encoding Approach to Store Image Data in DNA
DNA data storage is a promising technology that utilizes computer simulation and offers high-density and durable digital information storage. It is challenging to store massive image data in a small amount of DNA without losing the original data since nonspecific hybridization errors occur frequently and severely affect the durability of stored data. This work proposes a novel approach (RFS-codec) comprising an image fraction strategy and an innovative codec method to split and encode image data into DNA storage, respectively. The fraction strategy contributes by delivering a cost-effective solution for image storage in DNA. The codec method offers an encryption mechanism to convert binary data into DNA bases by avoiding hybridization errors and satisfying the critical bio-coding constraints responsible for DNA storage durability. The robustness of RFS-codec is computed with GC and homopolymer constraints. Experimentally, different image data are efficiently encoded and decoded successfully with 1.8 bit/nt average density. RFS-codec's results demonstrate substantial advantages in constructing cost-effective, scalable, and durable DNA data storage.
- 30 June 2025
- Published by Institute of Central Computation and Knowledge Inc. in Journal of Artificial Intelligence in Bioinformatics
- Vol. 1 (1) , 41-50
Open AccessRFS-codec: A Novel Encoding Approach to Store Image Data in DNA
DNA data storage is a promising technology that utilizes computer simulation and offers high-density and durable digital information storage. It is challenging to store massive image data in a small amount of DNA without losing the original data since nonspecific hybridization errors occur frequently and severely affect the durability of stored data. This work proposes a novel approach (RFS-codec) comprising an image fraction strategy and an innovative codec method to split and encode image data into DNA storage, respectively. The fraction strategy contributes by delivering a cost-effective solution for image storage in DNA. The codec method offers an encryption mechanism to convert binary data into DNA bases by avoiding hybridization errors and satisfying the critical bio-coding constraints responsible for DNA storage durability. The robustness of RFS-codec is computed with GC and homopolymer constraints. Experimentally, different image data are efficiently encoded and decoded successfully with 1.8 bit/nt average density. RFS-codec's results demonstrate substantial advantages in constructing cost-effective, scalable, and durable DNA data storage.
- Top Cited Papers
- 8 May 2025
- Published by Elsevier in Neurocomputing
- Vol. 641
Deep neurocomputational fusion for ASD diagnosis using multi-domain EEG analysis
- 8 May 2025
- Published by Elsevier in Neurocomputing
- Vol. 641
Top Cited PapersDeep neurocomputational fusion for ASD diagnosis using multi-domain EEG analysis
- Top Cited PapersOpen Access
- 9 April 2025
- Published by MDPI AG in Information
- Vol. 16 (4) , 301
nBERT: Harnessing NLP for Emotion Recognition in Psychotherapy to Transform Mental Health Care
The rising prevalence of mental health disorders, particularly depression, highlights the need for improved approaches in therapeutic interventions. Traditional psychotherapy relies on subjective assessments, which can vary across therapists and sessions, making it challenging to track emotional progression and therapy effectiveness objectively. Leveraging the advancements in Natural Language Processing (NLP) and domain-specific Large Language Models (LLMs), this study introduces nBERT, a fine-tuned Bidirectional Encoder Representations from the Transformers (BERT) model integrated with the NRC Emotion Lexicon, to elevate emotion recognition in psychotherapy transcripts. The goal of this study is to provide a computational framework that aids in identifying emotional patterns, tracking patient-therapist emotional alignment, and assessing therapy outcomes. Addressing the challenge of emotion classification in text-based therapy sessions, where non-verbal cues are absent, nBERT demonstrates its ability to extract nuanced emotional insights from unstructured textual data, providing a data-driven approach to enhance mental health assessments. Trained on a dataset of 2021 psychotherapy transcripts, the model achieves an average precision of 91.53%, significantly outperforming baseline models. This capability not only improves diagnostic accuracy but also supports the customization of therapeutic strategies. By automating the interpretation of complex emotional dynamics in psychotherapy, nBERT exemplifies the transformative potential of NLP and LLMs in revolutionizing mental health care. Beyond psychotherapy, the framework enables broader LLM applications in the life sciences, including personalized medicine and precision healthcare.
- 9 April 2025
- Published by MDPI AG in Information
- Vol. 16 (4) , 301
Top Cited PapersOpen AccessnBERT: Harnessing NLP for Emotion Recognition in Psychotherapy to Transform Mental Health Care
The rising prevalence of mental health disorders, particularly depression, highlights the need for improved approaches in therapeutic interventions. Traditional psychotherapy relies on subjective assessments, which can vary across therapists and sessions, making it challenging to track emotional progression and therapy effectiveness objectively. Leveraging the advancements in Natural Language Processing (NLP) and domain-specific Large Language Models (LLMs), this study introduces nBERT, a fine-tuned Bidirectional Encoder Representations from the Transformers (BERT) model integrated with the NRC Emotion Lexicon, to elevate emotion recognition in psychotherapy transcripts. The goal of this study is to provide a computational framework that aids in identifying emotional patterns, tracking patient-therapist emotional alignment, and assessing therapy outcomes. Addressing the challenge of emotion classification in text-based therapy sessions, where non-verbal cues are absent, nBERT demonstrates its ability to extract nuanced emotional insights from unstructured textual data, providing a data-driven approach to enhance mental health assessments. Trained on a dataset of 2021 psychotherapy transcripts, the model achieves an average precision of 91.53%, significantly outperforming baseline models. This capability not only improves diagnostic accuracy but also supports the customization of therapeutic strategies. By automating the interpretation of complex emotional dynamics in psychotherapy, nBERT exemplifies the transformative potential of NLP and LLMs in revolutionizing mental health care. Beyond psychotherapy, the framework enables broader LLM applications in the life sciences, including personalized medicine and precision healthcare.