translational medicine using graph-based problem oriented medical record (ql4pomr)

ql4pomr is a mitacs/nserc research project with thunder bay regional hospital

principle investigators: dr. sabah mohammed and dr. jinan fiaidhi

tbrhsc collaborator: dr. arnold kim (md)

ql4pomr is our translational medicine prototype representing a new vision to transform healthcare through interlinking everything from the bedside along with discoveries and knowledge available at the bench. it is uses the notion of graph (or the clinical problem list) to bridge both sides based on sound standards. ql4pomr brings several emerging methods to establish this bridge for clinical problem representation, clinical data integration, improving diagnosis, staging, prognosis, and treatment of diseases. ql4pomr integrates biomedical multi-omics data knowledge with the clinical practice data to improve the power of predicting clinical outcomes. since the electronic health records (ehr) are commonly used to store and analyze patient data, and then it seems straight-forward to perform research on modeling ehr data as graphs and bridge the bedside with the bench based on interlinked graphs. ql4pomr vision at the bedside starts with charting clinical cases according to soap note and building the clinical case graph based on common clinical problem list. this interlinked vision was originally introduced by lawrence weed (md) in his problem-oriented medical record (pomr) vision to solve interoperability in healthcare but was not implemented as the technologies of ehrs largely uses a tabular/relational format that is too different from the graph-based vision. ql4pomr based on the soap schema can translate and link clinical data to any ehr system or bench data repository server based on the graphql api. based on this flexible api ql4pomr managed to link clinical cases to the standard hl7 fhir electronic healthcare record as well as to produce the hl7 ips (international patient summary). based on the same api, ql4pomr interlinked patient cases and their records to the knowledge available at the bench. ql4pomr managed to connect to important biomedical repositories like the opentargets and drugbank. based on this graph-based vision, clinician at the bedside can see clearly alternative medications to what they have prescribed, they can see if their prescribed drugs can have some adverse interactions or if a patient is a case of a polypharmacy that requires deprescribing some of these medications. ql4pomer also help the bench side by providing an extended clinical trial verifications from the bedside. there are many other question that can be answered based on this graph-based translational vision to include the genetic data and the chemical components of the prescribed drugs. it is important to mention that ql4pomr is a result of my individual research as supported by recent nserc (2020-2023) and large collaboration with dr. fiaidhi and the lakehead graduate 世界杯2022赛程表淘汰赛 team as well as with thunder bay regional health science center (tbrhsc) through our 5 years mitacs project (2021-2025).

ql4pomr is the result of collaboration with thunder bay regional health science center

the mitacs research with tbrhsc is a five year project 2021-2025. it is a large research project that started in may 1st, 2021 to develop the next generation graph-based problem oriented medical record in collaboration with thunder bay regional health science center. i share this research with dr. jinan fiaidhi as principle investigator and dr. arnold kim (md) as our tbrhsc collaborator. this research has been also extended to cover remote patient monitoring and consultation through my new nserc discovery grant. both research uses my graph-based translational methods.

ql4pomr research objectives

there are so many objectives behind my cutting edge research. the first and most important objective is to bridge the gap between the bed-side clinical practice and the bench-side biomedical research where graph-based approaches can be used for the translational knowledge needed for building this bridge.

the second objective is the use of a standard graph-based schema for representing clinical cases at the bedside. ql4pomr uses the soap note schema (subjective as encountered by the patient, objectives as conducted by the examining physician observations, assessments from the requested lab and diagnostic imaging tests and the plan of treatment). the soap note was invented by dr. lary weed (md) and widely used in clinical education and practice. the third objective is to align the soap attributes to their corresponding medical problems and build a problem-oriented medical record (pomr) system that is centered on the notion of of the problem set. ql4pomr translates the soap/pomr schema into the standard hl7 fhir (resources) schema. the fourth objective is relate the pomr clinical description (via the fhir resources schema) to the corresponding biomedical knowledge available at the bench including data about drugs, diseases, genetics, adverse events, etc. through utilizing notable open source repositories like the opentargets and drugbank. the fifth objective is to employ the constructed bridge between the bedside and the bench side to provide focused analytics and to reveal important clinical patterns and links to benefit clinicians at the bedside (e.g. solving the ade effects of polypharmacy) and the biomedical/pharma scientists with a real world clinical monitoring (e.g. clinical trials from the bedside). the sixth objective is to use the contructed bridge between bedside and the bench to provide more evidence based analytics and precision knowledge. we are experimenting with using the new tranformer technologies like bert, biobert and gpt to produce sound clinical hypothesis from clinical questions using the pico standard both from physicians or from machine learning generated pico questions to interrigate and synthesis knowledge from medical repositories like pubmed.

to achieve all these objectives, this research project have constructed a prototype (ql4pomr) for the new problem oriented medical record (pomr) using the emerging graphql api to deal with semi structured care design and model the lower ends of soap as a graph. the early results of our research have demonstrated an amazing success to translate the soap cases into the standard hl7 fhir records based on the hl7 fhir server. the new ql4pomr system uses a crud (create, read, update, delete) interface to fetch and save semi-structured care data on the actual fhir system. we are adding more components to our ql4pomr to link it to other biomedical data through building a gatsby data layer that can incorporate schema of external data sources without having to have these data sources inside ql4pomr. the use of gatsby api through the graphql server prove very effective in linking to external biomedical sources such as opentargets and drugbank. more apis are used to generate the pico qestions and produce hypotheses summaries. our efforts is continuing and we are delighted that we have passionate group including our graduate 世界杯2022赛程表淘汰赛 .

the overall ql4pomr design:

the use of ql4pomr for evidence-based medicine

ql4pomr research timeline:

watch webinar on ql4pomr prototyping

https://www.youtube.com/watch?v=hwfgfppixnm

watch webinar on the ql4pomr pico wrapper:

https://www.youtube.com/watch?v=faa_e0czqiu

watch webinar on the ql4pomr evidence-based analytics:

https://www.youtube.com/watch?v=zxcg2kwwvv0

recent ql4pomr publications:

(1) establishment of a mindmap for medical e-diagnosis as a service for graph-based learning and analytics
s mohammed, j fiaidhi
springer neural computing & applications 33 (18), 1-12, 2021
https://link.springer.com/article/10.1007/s00521-021-06200-6

(2) ql4pomr interface as a graph-based clinical diagnosis web service
sabah mohammed, jinan fiaidhi, darien sawyer
ieee 11th international conference on logistics, informatics and service sciences (liss2021)
https://link.springer.com/chapter/10.1007/978-981-16-8656-6_5

(3) empowering graphql based problem oriented medical record systems using a data layer
sabah mohammed, jinan fiaidhi
international journal of future generation communication and networking (ijfgcn), volume 14, issue 3, pages 1-12,
http://article.nadiapub.com/ijfgcn/vol14_no3/1.html#

(4) generating physician standing orders for unplanned care scenarios using the hl7 fhir patient summaries,

sabah mohammed, jinan fiaidhi and darien sawyer, the 9th ieee international conference on e-health and bioengineering - ehb 2021 grigore t. popa university of medicine and pharmacy, web conference, romania, november 18-19, 2021.

https://ieeexplore.ieee.org/document/9657715

(5) problem oriented diagnostic service for describing clinical cases based on the graphql pomr approach,

sabah mohammed, jinan fiaidhi and darien sawyer, dec. 9-12, ieee bibm 2021
https://ieeexplore.ieee.org/abstract/document/9669364

(6) graphql patient case presentation using the problem oriented medical record schema,
sabah mohammed, jinan fiaidhi and darien sawyer, ieee big data, december 15-18, 2021
https://ieeexplore.ieee.org/abstract/document/9671394

(7) introducing ql4pomr crud bff for processing ips standard patient summary report on fhir

sabah mohammed; jinan fiaidhi; darien sawyer
4th ieee eurasia conference on biomedical engineering, healthcare and sustainability 2022 (ieee ecbios 2022)

https://www.ecbios.asia/

(8) value-based healthcare translational data analytics using the problem oriented medical record graph representation

sabah mohammed; jinan fiaidhi; darien sawyer
ichi 2022
rochester, minnesota, united states
11-14 june 2022

ichi 2022 conference proceedings: https://ieeexplore.ieee.org/abstract/document/9874641

(9) sabah mohammed; jinan fiaidhi; darien sawyer, problem oriented medical translational services based on graphql connectivity, ieee icc conference, seoul, korea, may 16-20, 2022:

ieee icc paper link: https://ieeexplore.ieee.org/abstract/document/9838418/

(10) investigating polypharmacy for patients with multi encounters using the ql4pomr framework, sabah mohammed, jinan fiaidhi and darien sawyer, ieee healthcom 2022, genoa, italy, oct 17-19, 2022

link to the ieee healthcom 2022 article: https://ieeexplore.ieee.org/abstract/document/9982796

(11) developing a graphql soap conversational micro frontends for the problem oriented medical record (ql4pomr), s. mohammed, j. fiaidhi, d. sawyer, m. lamouchie, acm icmhi 2022: 2022 6th international conference on medical and health informatics, kyoto, japan, may 13 - 15, 2022

links to the conference and the acm paper:

http://www.icmhi.org/icmhi2022.html

https://dl.acm.org/doi/abs/10.1145/3545729.3545738

(12) prototyping the problem oriented medical record for connected health based on typegraphql, sabah mohammed, jinan fiaidhi and darien sawyer, ieee big data 2022, dec. 17-20, 2022 osaka, japan.

link to the ieee big data 2022: https://bigdataieee.org/bigdata2022/

(13) problem-oriented medical records for describing care cases using multi-tenants
sabah mohammed, jinan fiaidhi, darien sawyer
ieee 2023 3rd international conference on innovative research in applied science, engineering and technology (iraset), morraco

link to ieee iraset 2023 paper: https://ieeexplore.ieee.org/abstract/document/10153018

(14)

problem-oriented translational health informatics for evidence based medicine and privacy enhancing
sabah mohammed and jinan fiaidhi, ieee big data servics 2023, july 17-20, 2023 athens, harokopio university of athens, greece.

link to the conference paper: https://ieeebigdataservice.com/accepted-papers/

(15) sabah mohammed and jinan fiaidhi, (2023), "investigation into scaling-up the soap problem-oriented medical record into a clinical case study", ieee ichi 2023, rice university, houston, taxes, usa, june 26-29, 2023

link to ieee ichi 2023: https://ieeeichi.github.io/ichi2023/program.html

(16) sabah mohammed and jinan fiaidhi, problem-oriented translational health informatics based on problem-oriented medical record, ieee 22nd international conference on cognitive informatics and cognitive computing (icci*cc 2023), standford university, august 19-21, 2023