Type:
Educational Exhibit
Keywords:
Management, Artificial Intelligence, Professional Issues, Professional issues, Experimental, PACS, Decision analysis, Efficacy studies, Structured reporting, Education and training, Workforce, Not applicable
Authors:
S. SIVARAMALINGAM; CHENNAI, TA/IN
DOI:
10.26044/ecr2020/C-14959
Findings and procedure details
THE FEATURES OF IIPP TABLE :
The IIPP table (Imaging Interpretation and Probability Percentage table) has the following features:
- We propose the table to be at the end of the report - after impression and before advice.
- The table has at least 4 columns and any ‘n’ number of rows according to the number of differentials involved. The columns are meant for serial number, pathological process, probability percentage and comments.
- The total probability percentage should be equal to hundred.
- The probability percentage is a representative of the radiologists' perspective only and is purely indented for better communication and reasoning; it is purely at the disposal of the radiologist.
- The table need not be used in direct findings leading to a single pathology. Eg: calculous cholecystitis.
ADVANTAGE OF THE IIPP TABLE:
- In case of limited resources, the clinician can direct further investigations first along the more probable diagnosis.
- Sometimes the invasive procedures warranted for the less probable diagnosis can always wait.
- Diseases with related pathological and radiological features can be clubbed in one row, for eg: terming tuberculosis and histoplasmosis under granulomatous infections (which cannot be distinguished radiologically in confidence).
SCOPE OF THE TABLE ON EMPOWERING THE RADIOLOGIST:
- This era started with BIRADS. BIRADS was not something that was easily accepted in the initial phases. It required a lot of commitment on the part of the radiologist to first study the ever-evolving guidelines and basic categorisation. From the side of clinicians, they had to educate themselves on the basic principles behind BIRADS reporting, before accepting the recommended line of management according to the scoring. But in the long run both the radiologists and the clinicians understood the advantage of using a tried and tested road-map, instead of treading unknown waters. The success of BIRADS system can be easily understood by the extension of similar systems into other arena like TIRADS, LIRADS, PIRADS etc.In a way, we can only imagine that these systems indirectly empower the radiologist to opine on the probable percentage of malignancy based on the imaging parameters. If we consider BIRADS -3 category, it allows the radiologist to suggest the probable but low likelihood of malignancy which is variable from 2-95%.
- But in routine reporting the radiologist is expected to give a single likely diagnosis based on the imaging findings. In atypical cases he/ she is expected to give a list of differentials, not bothered about what goes through his mind regarding his preferences.
- IIPP table can be considered similar to these scoring systems, for disease processes where strict guidelines have not been laid hitherto.
- In the long run, this increases the commitment of the radiologist towards giving strong opinions on cases, taking into considerations the clinical scenario, managements options etc.,. This will increase the gratification on the part of the radiologist and also will encourage the clinicians to interact more and more with the radiologist.
- This next step towards commitment can waive-off the scare of artificial intelligence in radiology. Come what may, without well-laid guidelines, the capability of the AI technology to come to a decision making is expected to be very limited.
STAND FROM LEGAL POINT OF VIEW:
It is to be understood that the decision making of a radiologist depends on various parameters ranging from the image acquisition protocol, the clinical information provided, the supportive blood investigations, patient co-operation during the study, etc.,
Hence it is unfair to make him/ her liable for commenting/ committing on a diagnosis inspite of certain uncertainties. So the IIPP table should ideally be exempted from being approached medico-legally. Otherwise the radiologists will be forced to safeguard themselves with non-committal terms and underplaying their skills.
FURTHER STUDIES/ DIRECTION:
Comparative / Validation studies can be done between Reports with Probability percentage table and those without, by obtaining scores on clarity, usefulness in guiding management- from clinicians to validate the advantage of this table.
The efficiency of interpretation in a radiology department can be improved by reviewing and analysing the discordant cases where there is disagreement in the final clinical/ HPE diagnosis with the probability percentage.