Healthcare narrative (such as clinical notes, discharge letters, nurse handover notes, imaging reports, patients posts on social media or feedback comments, etc.) has been used as a key communication stream that contains the majority of actionable and contextualised data, but which – despite being increasingly available in a digital form – is not routinely analysed, and is rarely integrated with other healthcare data on a large-scale. The are many barriers and challenges in processing healthcare free text, including, for example, the variability and implicit nature of language expressions, and difficulties in sharing training and evaluation data. On the other hand, recent years have witnessed increasing needs and opportunities to process free text, with a number of success stories that have demonstrate the feasibility of using advanced Natural Language Processing to unlock evidence contained in free text to support clinical care, patient self-management, epidemiological research and audit.
HealTAC 2018 will bring the academic, clinical, industrial and patient communities together to discuss the current state of the art in processing healthcare free text and share experience, results and challenges. We invite various types of contributions, including long and short papers, posters, PhD student papers, demos and panels, that address the variety of aspects involved in processing and using healthcare free text, including but not limited to
- Natural language processing of healthcare text
- Information extraction: identification of clinical variables and their values in free-text
- Medical ontologies and coding of healthcare text
- Machine-learning approaches to healthcare text analytics
- Transfer learning for healthcare text analytics
- Processing patient-generated data (e.g. social media, health forums, diaries)
- Processing clinical literature and trial reports
- Integration of structured and unstructured resources for health applications
- Text analytics and learning health systems
- Real-time processing of healthcare free text
- Real-world application of text analytics
- Implementation of healthcare text analytics in practice: public engagement and trust
- Sharing resources for healthcare text analytics (data and methods)
- Evaluation and assessment of text analytics methods
- Text mining for veterinary medicine
- Processing speech in healthcare applications
- Extended deadline for long and short papers and panel proposals: January 21st, 2018
- Deadline for posters, PhD forum and demo proposals: February 15th, 2018
- Notification of acceptance: February 22nd, 2018
All submissions should be submitted in the PDF format via EasyChar.
Submissions should describe original, completed and unpublished work that focuses on either methodological or applicational aspects. Full papers may consist of up to 8 pages; accepted papers will be orally presented at the workshop. Selected full papers will be invited for a special issue in a health informatics journal.
Submissions will describe original and unpublished work, potentially ongoing but with some findings; these will be presented as flash/lightning talks and will be also available as posters. Short papers may focus on either methodological or applicational aspects and may consist of up to 4 pages.
Posters will describe ideas or ongoing work (e.g. projects), and will be offered space for presentation during the conference. Posters should be single-page submissions.
PhD student papers
PhD student papers will be presented at the PhD forum, which will feature projects in healthcare text analytics. Contributions should consists of up to 4 pages and should present ongoing PhD research (in any stage). The forum will provide an opportunity for students to present their work to their peers and receive feedback from the community. These submissions will be treated separately from the other submissions. See more details.
During the conference, we will organise a number of special sessions, including:
Panel discussions can be proposed to address the main challenges in processing healthcare free-text or to discuss the future of particular methodologies. Panels will be allocated 1.5 hour slots for discussions. Proposals for panels should consists of up to 2 pages. See more details.
Demo sessions will provide a forum for demonstration of solutions and projects to the wider community. Proposals for demos should consists of up to 2 pages. Up to 30-minute slots will be made available. See more details.
For all paper contributions, please use HealTAC2018-Submission-Template.
For special session proposals, please use HealTAC2018-Session-Template.
All contributions should be submitted in PDF format to: EasyChair (https://easychair.org/conferences/?conf=healtac201).