Table 3.
Literature map of the included articles.
| References | Study type | Participation | Study duration | Intervention and/or methods | Outcomes |
|---|---|---|---|---|---|
| Cohn et al. (31) | Experimental study | 11 CD patients who underwent MRE between 2010 and 2015 | Not applicable to the type of study | CD patients who underwent MRE between 2010 and 2015 and for whom there was at least 6 months of follow-up or an outcome of interest, were retrospectively identified. An expert radiologist demarcated regions of interest (ROI) based on accepted MRE criteria | Radiomics-based ML analysis of MRE (resonance heterography) images of CD patients can be used to develop a personalized risk score to predict response to IS (immunosuppressive) therapy |
| Afzali et al. (15) | Case studies | 867 IBD patients were screened | Not applicable to the type of study | Data from EMRs extracted manually | The screen rate failed in 91.8% |
| Mossoto et al. (32) | Experimental study | 287 children with PIBD | Not applicable to the type of study | Mathematical model assembled different techniques of supervised ML to classify IBD diagnosis in patients with pediatrics | Clinical IBD potential of ML models |
| Roccetti et al., (33) | Longitudinal study | Crohn's disease experts with a specific pharmaceutical treatment, the infliximab | 2 yearsa | Participants postb were read and analyzed by human beings and automatic tools in order to understand their mood when infliximab treatment was mentioned (positives, neutral, or negative term exist in a given discourse) | Gastroenterologists tends to express more positive considerations than the OponionFinder. The non-medical experts tend to return a large number of negatives |
| Dardzinska and Kasperczuk, (16) | Experimental study | IBD patients | Not applicable to the type of study | The presented model predicts the probability of IBD with malignancy or benign tumors | Classification model tool to find symptoms that affect whether the patients is ill or notc |
| Ashton et al. (34)26 | Case studies | This study phase is not included in the inclusion of participants | Not applicable to the type of study | Literature review to the analysis of the current management of pediatric IBD applying personalized medicine | AI and ML for personalized medicined |
| Ashton and Beattie, (35) | Case studies | 400 patients in remission until 12 months | Model to predict disease outcome | Identify the potential to translate clinical data from diagnosis into a clinically accurate model predicting the response of medications, complications and others | |
| Borland et al. (36) | Experimental study | This study phase is not included in the inclusion of participants | Not applicable to the type of study | Integrative visualization tool enabling users to explore patients generated research questions or topics | |
| Zand et al. (37) | Experimental study | 1712 IBD patients | Do not have this information | Electronic dialog data collected between 2013 and 2018 from a care management platform (eIBD) at a tertiary referral center for IBD at the University of California, Los Angeles (UCLA) | Algorithm showed 94% similarity in categorization compared with our three independent physicians |
a
The 261 posts analysis range was between October 2013-October 2015.
b
Group A – Experts gastroenterologists; Group B – Non-medical experts; OpinionFinder – Standard values into the integer interval (−1,1) based on the algebraic sum of provided scores.
c
IBD doesn't present a precise diagnosis.
d
This is the only study founded that refer the potential of personalized medicine within crossing multi-omics data with clinical data (bloods, complications, outcomes, relapse, etc.).