Model Base Model Is Cased Train Datasets Inference speed (100 rows)
sbiobert_jsl_cased biobert_v1.1_pubmed Cased medNLI, allNLI 274,53
sbiobert_jsl_umls_cased biobert_v1.1_pubmed Cased medNLI, allNLI, umls 274,52
sbert_jsl_medium_uncased uncased_L-8_H-512_A-8 Uncased medNLI, allNLI 80,40
sbert_jsl_medium_umls_uncased uncased_L-8_H-512_A-8 Uncased medNLI, allNLI, umls 78,35
sbert_jsl_mini_uncased uncased_L-4_H-256_A-4 Uncased medNLI, allNLI 10,68
sbert_jsl_mini_umls_uncased uncased_L-4_H-256_A-4 Uncased medNLI, allNLI, umls 10,29
sbert_jsl_tiny_uncased uncased_L-2_H-128_A-2 Uncased medNLI, allNLI 4,54
sbert_jsl_tiny_umls_uncased uncased_L-2_H-128_A-2 Uncased medNLI, allNL, umls 4,54

Higher Accuracy ICD-10-CM Resolver Models

These models map clinical entities and concepts to ICD-10-CM codes using SBERT sentence embeddings. They also return the official resolution text within the brackets inside the metadata. Both models are augmented with synonyms, and previous augmentations are flexed according to cosine distances to unnormalized terms (ground truths).

  • sbiobertresolve_icd10cm_slim_billable_hcc: Trained with classic sbiobert mli. (sbiobert_base_cased_mli)

Models Hub Page:
https://nlp.johnsnowlabs.com/2021/05/25/sbiobertresolve_icd10cm_slim_billable_hcc_en.html

  • sbertresolve_icd10cm_slim_billable_hcc_med: Trained with new jsl sbert(sbert_jsl_medium_uncased)

Models Hub Page:
https://nlp.johnsnowlabs.com/2021/05/25/sbertresolve_icd10cm_slim_billable_hcc_med_en.html

Example: ‘bladder cancer’

  • sbiobertresolve_icd10cm_augmented_billable_hcc
chunks code all_codes resolutions all_distances 100x Loop(sec)
bladder cancer C679 [C679, Z126, D090, D494, C7911] [bladder cancer, suspected bladder cancer, cancer in situ of
urinary bladder, tumor of bladder neck, malignant tumour of
bladder neck]
[0.0000, 0.0904, 0.0978, 0.1080, 0.1281] 26,9
  • sbiobertresolve_icd10cm_slim_billable_hcc
chunks code all_codes resolutions all_distances 100x Loop(sec)
bladder cancer D090 [D090, D494, C7911, C680, C679] [cancer in situ of urinary bladder [Carcinoma in situ of bladder],
tumor of bladder neck [Neoplasm of unspecified behavior of
bladder], malignant tumour of bladder neck [Secondary malignant
neoplasm of bladder], carcinoma of urethra [Malignant neoplasm
of urethra], malignant tumor of urinary bladder [Malignant neoplasm
of bladder, unspecified]]
[0.0978, 0.1080, 0.1281, 0.1314, 0.1284] 20,9
  • sbertresolve_icd10cm_slim_billable_hcc_med
chunks code all_codes resolutions all_distances 100x Loop(sec)
bladder cancer C671 [C671, C679, C61, C672, C673] [bladder cancer, dome [Malignant neoplasm of dome of bladder], cancer of the
urinary bladder [Malignant neoplasm of bladder, unspecified], prostate cancer
[Malignant neoplasm of prostate], cancer of the urinary bladder]
[0.0894, 0.1051, 0.1184, 0.1180, 0.1200] 12,8