A Hybrid Ensemble Learning Framework for Predicting Lumbar Disc Herniation Recurrence: Integrating Supervised Models, Anomaly Detection, and Threshold Optimization
<b>Background:</b> Lumbar disc herniation (LDH) recurrence remains a pressing clinical challenge, with limited predictive tools available to support early identification and personalized intervention. Predicting recurrence after lumbar disc herniation (LDH) remains clinically important b...
Saved in:
| Main Authors: | , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2025-06-01
|
| Series: | Diagnostics |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2075-4418/15/13/1628 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1849428693757198336 |
|---|---|
| author | Mădălina Duceac (Covrig) Călin Gheorghe Buzea Alina Pleșea-Condratovici Lucian Eva Letiția Doina Duceac Marius Gabriel Dabija Bogdan Costăchescu Eva Maria Elkan Cristian Guțu Doina Carina Voinescu |
| author_facet | Mădălina Duceac (Covrig) Călin Gheorghe Buzea Alina Pleșea-Condratovici Lucian Eva Letiția Doina Duceac Marius Gabriel Dabija Bogdan Costăchescu Eva Maria Elkan Cristian Guțu Doina Carina Voinescu |
| author_sort | Mădălina Duceac (Covrig) |
| collection | DOAJ |
| description | <b>Background:</b> Lumbar disc herniation (LDH) recurrence remains a pressing clinical challenge, with limited predictive tools available to support early identification and personalized intervention. Predicting recurrence after lumbar disc herniation (LDH) remains clinically important but algorithmically difficult due to extreme class imbalance and low signal-to-noise ratio. <b>Objective:</b> This study proposes a hybrid machine learning framework that integrates supervised classifiers, unsupervised anomaly detection, and decision threshold tuning to predict LDH recurrence using routine clinical data. <b>Methods:</b> A dataset of 977 patients from a Romanian neurosurgical center was used. We trained a deep neural network, random forest, and an autoencoder (trained only on non-recurrence cases) to model baseline and anomalous patterns. Their outputs were stacked into a meta-classifier and optimized via sensitivity-focused threshold tuning. Evaluation was performed via stratified cross-validation and external holdout testing. <b>Results:</b> Baseline models achieved high accuracy but failed to recall recurrence cases (0% sensitivity). The proposed ensemble reached 100% recall internally with a threshold of 0.05. Key predictors included hospital stay duration, L4–L5 herniation, obesity, and hypertension. However, external holdout performance dropped to 0% recall, revealing poor generalization. <b>Conclusions:</b> The ensemble approach enhances detection of rare recurrence cases under internal validation but exhibits poor external performance, emphasizing the challenge of rare-event modeling in clinical datasets. Future work should prioritize external validation, longitudinal modeling, and interpretability to ensure clinical adoption. |
| format | Article |
| id | doaj-art-4cce1ed9811a4dca968f0cd60b66c180 |
| institution | Kabale University |
| issn | 2075-4418 |
| language | English |
| publishDate | 2025-06-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Diagnostics |
| spelling | doaj-art-4cce1ed9811a4dca968f0cd60b66c1802025-08-20T03:28:37ZengMDPI AGDiagnostics2075-44182025-06-011513162810.3390/diagnostics15131628A Hybrid Ensemble Learning Framework for Predicting Lumbar Disc Herniation Recurrence: Integrating Supervised Models, Anomaly Detection, and Threshold OptimizationMădălina Duceac (Covrig)0Călin Gheorghe Buzea1Alina Pleșea-Condratovici2Lucian Eva3Letiția Doina Duceac4Marius Gabriel Dabija5Bogdan Costăchescu6Eva Maria Elkan7Cristian Guțu8Doina Carina Voinescu9Faculty of Medicine and Pharmacy, Doctoral School of Biomedical Sciences, “Dunărea de Jos” University of Galați, 47 Domnească Street, 800008 Galați, RomaniaClinical Emergency Hospital “Prof. Dr. Nicolae Oblu”, 700309 Iași, RomaniaFaculty of Medicine and Pharmacy, “Dunărea de Jos” University of Galați, 47 Domnească Street, RO-800008 Galați, RomaniaClinical Emergency Hospital “Prof. Dr. Nicolae Oblu”, 700309 Iași, RomaniaClinical Emergency Hospital “Prof. Dr. Nicolae Oblu”, 700309 Iași, RomaniaClinical Emergency Hospital “Prof. Dr. Nicolae Oblu”, 700309 Iași, RomaniaClinical Emergency Hospital “Prof. Dr. Nicolae Oblu”, 700309 Iași, RomaniaFaculty of Medicine and Pharmacy, “Dunărea de Jos” University of Galați, 47 Domnească Street, RO-800008 Galați, RomaniaFaculty of Medicine and Pharmacy, “Dunărea de Jos” University of Galați, 47 Domnească Street, RO-800008 Galați, RomaniaFaculty of Medicine and Pharmacy, “Dunărea de Jos” University of Galați, 47 Domnească Street, RO-800008 Galați, Romania<b>Background:</b> Lumbar disc herniation (LDH) recurrence remains a pressing clinical challenge, with limited predictive tools available to support early identification and personalized intervention. Predicting recurrence after lumbar disc herniation (LDH) remains clinically important but algorithmically difficult due to extreme class imbalance and low signal-to-noise ratio. <b>Objective:</b> This study proposes a hybrid machine learning framework that integrates supervised classifiers, unsupervised anomaly detection, and decision threshold tuning to predict LDH recurrence using routine clinical data. <b>Methods:</b> A dataset of 977 patients from a Romanian neurosurgical center was used. We trained a deep neural network, random forest, and an autoencoder (trained only on non-recurrence cases) to model baseline and anomalous patterns. Their outputs were stacked into a meta-classifier and optimized via sensitivity-focused threshold tuning. Evaluation was performed via stratified cross-validation and external holdout testing. <b>Results:</b> Baseline models achieved high accuracy but failed to recall recurrence cases (0% sensitivity). The proposed ensemble reached 100% recall internally with a threshold of 0.05. Key predictors included hospital stay duration, L4–L5 herniation, obesity, and hypertension. However, external holdout performance dropped to 0% recall, revealing poor generalization. <b>Conclusions:</b> The ensemble approach enhances detection of rare recurrence cases under internal validation but exhibits poor external performance, emphasizing the challenge of rare-event modeling in clinical datasets. Future work should prioritize external validation, longitudinal modeling, and interpretability to ensure clinical adoption.https://www.mdpi.com/2075-4418/15/13/1628lumbar disc herniationrecurrence predictionensemble learningautoencoderthreshold tuningclass imbalance |
| spellingShingle | Mădălina Duceac (Covrig) Călin Gheorghe Buzea Alina Pleșea-Condratovici Lucian Eva Letiția Doina Duceac Marius Gabriel Dabija Bogdan Costăchescu Eva Maria Elkan Cristian Guțu Doina Carina Voinescu A Hybrid Ensemble Learning Framework for Predicting Lumbar Disc Herniation Recurrence: Integrating Supervised Models, Anomaly Detection, and Threshold Optimization Diagnostics lumbar disc herniation recurrence prediction ensemble learning autoencoder threshold tuning class imbalance |
| title | A Hybrid Ensemble Learning Framework for Predicting Lumbar Disc Herniation Recurrence: Integrating Supervised Models, Anomaly Detection, and Threshold Optimization |
| title_full | A Hybrid Ensemble Learning Framework for Predicting Lumbar Disc Herniation Recurrence: Integrating Supervised Models, Anomaly Detection, and Threshold Optimization |
| title_fullStr | A Hybrid Ensemble Learning Framework for Predicting Lumbar Disc Herniation Recurrence: Integrating Supervised Models, Anomaly Detection, and Threshold Optimization |
| title_full_unstemmed | A Hybrid Ensemble Learning Framework for Predicting Lumbar Disc Herniation Recurrence: Integrating Supervised Models, Anomaly Detection, and Threshold Optimization |
| title_short | A Hybrid Ensemble Learning Framework for Predicting Lumbar Disc Herniation Recurrence: Integrating Supervised Models, Anomaly Detection, and Threshold Optimization |
| title_sort | hybrid ensemble learning framework for predicting lumbar disc herniation recurrence integrating supervised models anomaly detection and threshold optimization |
| topic | lumbar disc herniation recurrence prediction ensemble learning autoencoder threshold tuning class imbalance |
| url | https://www.mdpi.com/2075-4418/15/13/1628 |
| work_keys_str_mv | AT madalinaduceaccovrig ahybridensemblelearningframeworkforpredictinglumbardischerniationrecurrenceintegratingsupervisedmodelsanomalydetectionandthresholdoptimization AT calingheorghebuzea ahybridensemblelearningframeworkforpredictinglumbardischerniationrecurrenceintegratingsupervisedmodelsanomalydetectionandthresholdoptimization AT alinapleseacondratovici ahybridensemblelearningframeworkforpredictinglumbardischerniationrecurrenceintegratingsupervisedmodelsanomalydetectionandthresholdoptimization AT lucianeva ahybridensemblelearningframeworkforpredictinglumbardischerniationrecurrenceintegratingsupervisedmodelsanomalydetectionandthresholdoptimization AT letitiadoinaduceac ahybridensemblelearningframeworkforpredictinglumbardischerniationrecurrenceintegratingsupervisedmodelsanomalydetectionandthresholdoptimization AT mariusgabrieldabija ahybridensemblelearningframeworkforpredictinglumbardischerniationrecurrenceintegratingsupervisedmodelsanomalydetectionandthresholdoptimization AT bogdancostachescu ahybridensemblelearningframeworkforpredictinglumbardischerniationrecurrenceintegratingsupervisedmodelsanomalydetectionandthresholdoptimization AT evamariaelkan ahybridensemblelearningframeworkforpredictinglumbardischerniationrecurrenceintegratingsupervisedmodelsanomalydetectionandthresholdoptimization AT cristiangutu ahybridensemblelearningframeworkforpredictinglumbardischerniationrecurrenceintegratingsupervisedmodelsanomalydetectionandthresholdoptimization AT doinacarinavoinescu ahybridensemblelearningframeworkforpredictinglumbardischerniationrecurrenceintegratingsupervisedmodelsanomalydetectionandthresholdoptimization AT madalinaduceaccovrig hybridensemblelearningframeworkforpredictinglumbardischerniationrecurrenceintegratingsupervisedmodelsanomalydetectionandthresholdoptimization AT calingheorghebuzea hybridensemblelearningframeworkforpredictinglumbardischerniationrecurrenceintegratingsupervisedmodelsanomalydetectionandthresholdoptimization AT alinapleseacondratovici hybridensemblelearningframeworkforpredictinglumbardischerniationrecurrenceintegratingsupervisedmodelsanomalydetectionandthresholdoptimization AT lucianeva hybridensemblelearningframeworkforpredictinglumbardischerniationrecurrenceintegratingsupervisedmodelsanomalydetectionandthresholdoptimization AT letitiadoinaduceac hybridensemblelearningframeworkforpredictinglumbardischerniationrecurrenceintegratingsupervisedmodelsanomalydetectionandthresholdoptimization AT mariusgabrieldabija hybridensemblelearningframeworkforpredictinglumbardischerniationrecurrenceintegratingsupervisedmodelsanomalydetectionandthresholdoptimization AT bogdancostachescu hybridensemblelearningframeworkforpredictinglumbardischerniationrecurrenceintegratingsupervisedmodelsanomalydetectionandthresholdoptimization AT evamariaelkan hybridensemblelearningframeworkforpredictinglumbardischerniationrecurrenceintegratingsupervisedmodelsanomalydetectionandthresholdoptimization AT cristiangutu hybridensemblelearningframeworkforpredictinglumbardischerniationrecurrenceintegratingsupervisedmodelsanomalydetectionandthresholdoptimization AT doinacarinavoinescu hybridensemblelearningframeworkforpredictinglumbardischerniationrecurrenceintegratingsupervisedmodelsanomalydetectionandthresholdoptimization |