The amount of data generated by the Healthcare industry over the years is enormous. A vast amount of this information comes from unstructured data in the form of physician diagnosis. Although the EHRs promise a lot, the problem is it’s time-consuming, while the same can be used to treat the patient. To tackle this issue a new logical approach that has been emerging in the world of Data science today is Natural Language Processing. It offers a wide spectrum of automated tools making the whole user experience much more seamless. To add more to it such NLP algorithms have been developed that have proven to be very useful for the healthcare industry, doctors, and patients. NLP in healthcare offers a wide spectrum of possibilities which are otherwise limited.
Dissecting it further
Clinicians have been using computerized clinical decision support systems (CDSS) for decades now. These tools mostly include Electronic Health Reports (EHRs). This includes EHRs that guide disease management, calculate drug dosing, or remind clinicians of patient-specific drug interactions and allergies.
Although EHRs hold a lot of information large portions of it remain unstructured in Narrative forms of patient observations over a period of time. This opens up the opportunity for new technology.
NLP in healthcare offers a wide spectrum of tools that can help clinicians process the data faster. One such NLP algorithm was trained. This NLP tool was trained on 22 emergency department summaries of patients who were later diagnosed with KD, and when applied to 253 emergency department notes for children who were diagnosed with either KD or another febrile illness, the tool identified patients of high suspicion for KD with 93.6% sensitivity and 77.5% specificity. For rare conditions with serious complications, such as KD, automated NLP-enabled computer alerts about suspicious presentations could prove to be life-saving and cost-effective.
NLP in healthcare offers a precise methodological premise for medical research. A qualitative approach, such as a thematic analysis of patient interviews, can offer more data and unexpected insights than a quantitative approach such as a simple Likert scale-based survey. Indeed, in a study where physicians were asked to rate two fictitious abstracts with the same research topic and title, differing only by whether the design was quantitative or qualitative, primary care physicians appreciated the qualitative study as more clinically relevant.
Coming to the end user ,
Tele-transportation services such as Telehealth Ontario provide telephone access to nurses who can provide basic disease information and care instructions. Unfortunately, this service is quite expensive; In Ontario, increasing program costs and lowering total call volume resulted in a price tag of $ 43. An innovative solution that could reduce the amount of resources invested in tele-triage is the use of an NLP-enabled chatbot that allows dialogue to speak or write about patient health issues. Although we could not find a description of the chatbot supporting NLP at this time in the health care sector, the NLP tool has been used successfully in clinical text analysis, which suggests that an NLP-enabled chatbot can process health-related texts. Recently such chatbots have been successfully implemented by companies such as Lufthansa, BMW, and Panasonic to provide customer and technical support. While customer support healthcare varies widely, users can learn important lessons for developing chatbots with NLP industry capabilities.
A tele-triage chatbot would free up the nurses and physicians who currently provide and monitor such services, who could otherwise apply their expertise in a face-to-face setting with patients.
So what might limit it?
Legal and ethical issues associated with artificial intelligence services such as relocation, accountability, and confidentiality are essential for assessing the potential use of NLP in health care. Although NLP can improve quality and access to health care, steps need to be taken to implement NLP safely and ethically.
Although a number of studies have demonstrated cases where NLP has been able to use qualitative information to predict outcomes, many algorithms currently require further development before they can be considered for use in CDSS. A recent study looked at suicidal ideation in a textual mental health analysis and gave only 61% and 56% sensitivity and specificity, respectively. There is also concern about developing an NLP algorithm to increase physician bias. The predictable accuracy of the algorithm depends on the quality of the data entered. When these algorithms use data generated by doctors, and the process for collecting that data has inherent variants, they will reflect these underlying assumptions and assumptions. Speech input based NLP device development is also based on the UWOMJ Fall 2018 article on speech-to-text translation accuracy. Simply put, NLP with garbage input makes its way out of the trash. After all, the NLP still promises to advance the healthcare industry. As algorithms become stronger and more sophisticated, the widespread use of NLP will allow patients and health care professionals to benefit from a wide range of data sources, thereby increasing the quality and efficiency of health care.
So the word is,
Traditional tools for assessing customer satisfaction are limited to quantitative surveys. However, these surveys cannot cover the entire spectrum of user experiences and are subject to systematic bias. Much qualitative information that could be used effectively was also left in the questionnaire. NLP has the ability to receive patient stories in the form of unstructured and unrestricted feedback, both written and oral, and effectively produce meaningful summaries for the management team. This will provide new tools and insights that will assist in the implementation and monitoring of quality improvement initiatives.
Halper F. Advanced analytics: moving toward AI, machine learning,
and natural language processing. Available from: https://www.sas.com/content/dam/SAS/en_us/doc/whitepaper2/tdwi-advanced-analytics-ai-ml-nlp-109090.pdf.
Crowston K, Allen EE, Heckman R. Using natural language processing
technology for qualitative data analysis. Int J Soc Res Methodol. 2012
Tierney PJ. A qualitative analysis framework using natural language
processing and graph theory. Int Rev Res Open Distrib Learn Distrib.