Tieto­evry Care — a dig­i­tal assis­tant for care

Many coun­tries face over­bur­dened health and social care sys­tems, with aging pop­u­la­tions dri­ving an increased need for care. Pub­lic health­care bud­gets are not grow­ing at the same rate, and the avail­abil­i­ty of skilled work­ers may also be con­strained. Thus, the health sec­tor has a high need for tools that can help pro­vide cost-effec­tive and high-qual­i­ty care. Machine learn­ing and arti­fi­cial intel­li­gence are crit­i­cal in enabling this. Tieto­evry Care is on a mis­sion to reduce the bur­den on care work­ers and improve health out­comes by using AI. The com­pa­ny has been work­ing with AI in health­care for sev­er­al years and has mul­ti­ple AI projects under­way. This has helped Tieto­evry Care learn a lot about cus­tomer needs, tech­nol­o­gy chal­lenges, and asse­ing the suit­abil­i­ty of AI.

“We envi­sion AI being a health and social care assis­tant, with care pro­fes­sion­als able to use the tech­nol­o­gy to sur­face, e.g., valu­able infor­ma­tion, write up patient notes, and even make con­nec­tions between var­i­ous data points. Automat­ing admin­is­tra­tive tasks in this way frees up more time car­ing for patients,” explained Niina Siipo­la, Head of AI and Data Solu­tions at Tieto­evry Care.

Tietoevry’s Life­care Data Plat­form enables the mul­ti-source data col­lec­tion and man­age­ment need­ed to pow­er AI-dri­ven health and social care ser­vices. The plat­form is wide­ly used in Fin­land. A key part of the offer­ing is a solu­tion that removes any poten­tial­ly iden­ti­fy­ing data before Large Lan­guage Mod­els (LLMs) are used. This capa­bil­i­ty is not a giv­en among soft­ware providers work­ing with health­care data.

One of the reg­u­la­to­ry devel­op­ments guid­ing Tieto­evry Care’s work in Fin­land is a legal change that allows med­ical pro­fes­sion­als to proac­tive­ly con­tact cit­i­zens. This is rel­e­vant when machine learn­ing tools pick up on dis­ease mark­ers that humans may over­look. The author­i­ties can now con­tact a per­son who has a clin­i­cal­ly sig­nif­i­cant find­ing for a spe­cif­ic dis­ease unless that per­son has opt­ed out of being con­tact­ed. Tieto­evry Care and Helsin­ki Uni­ver­si­ty Hos­pi­tal (HUS) (fea­tured in this CNN health­care report) have been work­ing togeth­er in this domain, devel­op­ing the machine learn­ing algo­rithms and data lake capa­bil­i­ties need­ed to diag­nose three groups of rare dis­eases.

Tieto­evry Care par­tic­i­pates in a new two-year research project to assess the poten­tial of LLMs for dif­fer­ent health­care use cas­es. The project, which is led by Helsin­ki Uni­ver­si­ty, received recent­ly more than one mil­lion euros in fund­ing. It brings togeth­er sev­er­al health­care orga­ni­za­tions and pri­vate com­pa­nies. An impor­tant use case being explored is how AI can speed up the doc­u­men­ta­tion process in both health­care and social care set­tings. Tieto­evry Care is cur­rent­ly pilot­ing a new solu­tion that uses AI speech-to-text tech­nol­o­gy, enabling care work­ers to dic­tate notes after patient vis­its instead of man­u­al­ly typ­ing them up. The solu­tion also struc­tures the data and cor­rects typos. Pow­ered by Microsoft Azure’s AI capa­bil­i­ties, the solu­tion is cur­rent­ly being test­ed by over 50 care pro­fes­sion­als to ensure it meets a wide range of needs.

“Finnish is prob­a­bly one of the most chal­leng­ing lan­guages for LLMs, espe­cial­ly in the clin­i­cal con­text. But we believe this pilot solu­tion can eas­i­ly be adapt­ed for the Swedish and Nor­we­gian mar­kets, where the lan­guage mod­els are an even bet­ter fit for the need,” said Siipo­la.

Read the whole arti­cle on Tieto­evry’s web­site.

Source: Tieto­evry