{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/joint-classification-and-prediction-cnn","title":"Joint Classification and Prediction CNN Framework for Automatic Sleep Stage Classification","arxiv_id":"1805.06546","date":"2018-05-16","proceeding":null,"authors":["Huy Phan","Fernando Andreotti","Navin Cooray","Oliver Y. Chén","Maarten De Vos"],"abstract":"Correctly identifying sleep stages is important in diagnosing and treating\nsleep disorders. This work proposes a joint classification-and-prediction\nframework based on CNNs for automatic sleep staging, and, subsequently,\nintroduces a simple yet efficient CNN architecture to power the framework.\nGiven a single input epoch, the novel framework jointly determines its label\n(classification) and its neighboring epochs' labels (prediction) in the\ncontextual output. While the proposed framework is orthogonal to the widely\nadopted classification schemes, which take one or multiple epochs as contextual\ninputs and produce a single classification decision on the target epoch, we\ndemonstrate its advantages in several ways. First, it leverages the dependency\namong consecutive sleep epochs while surpassing the problems experienced with\nthe common classification schemes. Second, even with a single model, the\nframework has the capacity to produce multiple decisions, which are essential\nin obtaining a good performance as in ensemble-of-models methods, with very\nlittle induced computational overhead. Probabilistic aggregation techniques are\nthen proposed to leverage the availability of multiple decisions. We conducted\nexperiments on two public datasets: Sleep-EDF Expanded with 20 subjects, and\nMontreal Archive of Sleep Studies dataset with 200 subjects. The proposed\nframework yields an overall classification accuracy of 82.3% and 83.6%,\nrespectively. We also show that the proposed framework not only is superior to\nthe baselines based on the common classification schemes but also outperforms\nexisting deep-learning approaches. To our knowledge, this is the first work\ngoing beyond the standard single-output classification to consider multitask\nneural networks for automatic sleep staging. This framework provides avenues\nfor further studies of different neural-network architectures for automatic\nsleep staging.","url_abs":"http://arxiv.org/abs/1805.06546v3","url_pdf":"http://arxiv.org/pdf/1805.06546v3.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"joint-classification-and-prediction-cnn","repo_url":"https://github.com/pquochuy/MultitaskSleepNet","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"tf","reach":null}],"tasks":[{"task_slug":"automatic-sleep-stage-classification","task_name":"Automatic Sleep Stage Classification"},{"task_slug":"classification-1","task_name":"Classification"},{"task_slug":"classification","task_name":"General Classification"},{"task_slug":"sleep-stage-detection","task_name":"Sleep Stage Detection"},{"task_slug":"sleep-staging","task_name":"Sleep Staging"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/sleep-stage-detection-on-mass-ss2","task":"Sleep Stage Detection","dataset":"MASS SS2","model":"Multitask\n1-max CNN","rank_in_archive_order":2,"of":2,"metrics":{"Accuracy":"78.6%"},"uses_additional_data":false},{"leaderboard":"/sota/sleep-stage-detection-on-sleep-edf","task":"Sleep Stage Detection","dataset":"Sleep-EDF","model":"Multitask 1-max CNN","rank_in_archive_order":7,"of":8,"metrics":{"Accuracy":"81.9%"},"uses_additional_data":true}],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}