Version: BR14.6🔗

Release date: 2025-03-13

Features/General🔗

• Parquet files generated for:
• Derived categories
  • Basic Demographics
  • Visit Data
• BioSpecimen
  • Nails
  • Urine
• Other
  • TLFB
  • SPM-2
  • MAP-DB
  • MLDS
  • Infant Sensor Questionnaire

MRI/EEG/BioSensors🔗

• MRI: Added Shadow Matrix files for MRI & EEG derivatives
• EEG: Removed 11 participants from EEG Made & QC
• BioSpecimen toxicology results: Added 'Candidate age' & 'Mother age'

Basic Demographics (sed_basic_demographics)🔗

• Set 'Unknown' for selected Basic Demographics fields at V01

Data Dictionary🔗

• Added Keys & Values for multi-select to checkbox fields (previously had no 'value' for 'levels')
• Set 'source' to 'General' and 'domain' to 'Demographics' for Basic Demographics and Visit Data categories
• Updated DD elements for MRI & EEG Derivatives (reviewed by WG)
  • type_data
  • type_level
  • type_var
• Updated DD elements for categorical fields
  • 'type_data': Set to 'Character' for 'dropdown' and selected MLDS 'text' fields
• Updated DD elements for 'Date' fields
  • 'type_data': Set to 'date'
  • 'type_level': Set to 'interval'
  • 'type_field': Set to 'text'
• Updated DD elements for numeric fields (as per definition in Types of scales & levels of measurement)
  • 'type_data': Set to 'double' for fields with decimal notation
  • 'type_data': Set to 'integer' for fields with no decimal notation
  • 'type_level': Set to 'interval' for fields with equal distance between values
  • 'type_level': Set to 'ratio' for fields with natural '0'
• Update DD element for score and derived fields
  • 'type_field': Left empty (no value set)
• Updated DD elements for ID/Interpretative score/Derived data/Medication names/Code fields
  • 'type_data': Set to 'text'
  • 'type_level': Set to 'nominal'
  • 'type_field': Set to 'text'
• Updated 'type_var' DD element designations for fields ('administrative'/'item'/'summary score')

Database/Scoring🔗

• 'Pex Health Pregnancy' medication labels ('pex_bm_health_preg__meds_003__01') initially entered as strings converted to RxNorm code (n=5)
• Removed entries with HTML text in 'healthv1 pex_bm_health_preg_meds' (003_02; 015__02)

Filters🔗

• Filter criteria:
  • Updated filter to not remove participants if 'Examiner' is 'Null' and 'Adminsitration' is 'None'
• Fields added:
  • BioSpecimens: Candidate age, Mother age
• Fields excluded:
  • Excluded TLFB 'filename' field
  • Remove 'gestational_age_at_birth' field (replaced with GAD)
  • Selected Demographics (sed_bm_demo) fields removed
  • Removed 'Descriptive' fields - Not used for data entry (n=228)

Counts🔗

• Total number of participants queried from DB: 1,443
• Total number of visits:
  • All: 2,232
  • V01: 1,443
  • V02: 666
  • V03: 123

BR14.6 Resolved Known Issues🔗

⚠️ pex_bm_health_preg Leaked HTML Code🔗

The field pex_bm_health_preg__meds_pex_bm_health_preg__meds_003__02 has HTML leak text in the output: <span style="color:#888;">[<span style="margin:0 2px;">No results were returned</span>]</span>.

⚠️ pex_bm_health_preg__meds_003__01 Medication Code Entries Replaced With GUID (Bug In De-Identitification Worfklow)🔗

The field pex_bm_health_preg__meds_003__01 has 3 entries that start with sub-XXXXXX instead of medication codes. Per Santiago, these entries are the GUID used to de-identify the data. This is due to the de-identification process, which recognizes the code for the medication as if it were a 'candID' and replaces it with the GUID (in this case the following codes were replaced with the corresponding GUID: 211965, 372596, 998033).

This is something that should be addressed on the next de-identification process. Harshitha is aware and will modify the de-identification process to run solely on the identifier columns (candID) and not on the data, as 6 digit numbers may be mistaken for IDs and replaced with GUIDs.

In addition, 5 entries that were incorrectly entered as strings instead of codes for this field were caught and replaced.

⚠️ Codes inserted as labels for 2 instances in pex_bm_health_preg__illness🔗

⚠️ Deleted EEG Data Causing 404 Errors🔗

The V03 acquisitions for 5 participants errored in MADE and had various issues, therefore the EEG Workgroup asked for those files to be removed from the release. However, the EEG files were removed from BR14 and BR14.5 post-ingestion, which lead to 404 file not found errors when users queried file-based data in BR14. Reloading BR14.5 has removed the indexes and resolved this issue (i.e. these missing files are no longer queriable via Lasso). There are still some EEG files available for those participants, but only files of format TSV and JSON for the channels and no main SET or EDF files.

⚠️ Globus Defaults to User's Home Directory on Windows🔗

For Mac users, Globus displays all the directories available on their system, but for Windows, it defaults to the path to the user's home directory. This is a known bug. Robust error handling will be added to handle paths that either Globus Connect doesn't have access to or paths that don't exist in the user's file system.

⚠️ Missing Labels For 'Levels' On Multi-Select To Checkbox Items🔗

This is due to an adjustment done to multi-select fields, whereby fields with multi-select options are converted into separate columns/fields as 'checkboxes'. In these cases, the values for the checkbox 'levels' is missing, and only the expected labels (TRUE/FALSE) are available in the DD.

Impacted instruments/fields:

• Multilingual Language Development Screener (ncl_ch_mlds)
• Infant Sensor Questionnaire 1 (nt_ch_sens__qtn_1)
• Infant Sensor Questionnaire 2 (nt_ch_sens__qtn_2)
• Infant Sensor Questionnaire 3 (nt_ch_sens__qtn_3)
• Pex Bm Health Preg I Chroncond (pex_bm_health_preg__chroncond)
• Pex Bm Health Preg I Erhosp (pex_bm_health_preg__erhosp)
• Pex Bm Health Preg I Exp I Vacc (pex_bm_health_preg__exp__vacc)
• Pex Bm Health Preg I Healthhx (pex_bm_health_preg__healthhx)
• Pex Bm Health Preg I Illness (pex_bm_health_preg__illness)
• Pex Bm Healthv2 Inf (pex_bm_healthv2_inf)
• Pex Bm Healthv2 Pregg (pex_bm_healthv2_preg)
• Height/Weight/Head Circumference (ph_ch_anthro)
• Basic Demographics (sed_basic_demographics)
• BFY - Benefits/Services, Economic Stress (sed_bm_bfy)
• Demographics (sed_bm_demo)
• Demographics (sed_bm_phx__discr)

⚠️ Missing Labels/Description For All Instruments🔗

To make the DD BIDS compliant, the change in the name of the DD metadata for the field description was changed from 'label' to 'Description' and the 'unit' to 'Units'. This change was not initially recognized by Lasso, causing the description for fields to be missing for all instruments.

⚠️ Empty Label Column in HBCD Data Dictionary🔗

The field dictionary "label" has been renamed "Description," diverging from ABCD convention. To fix this, Lasso will change the UI to read “DESCRIPTION” (instead of label); so wherever the user goes they will see the same thing (JSON or DD). Lasso will populate DESCRIPTION with DESCRIPTION in the JSON from HBCD (Lasso used to look for “LABEL” in the JSON, which is no longer present). For ABCD: “‘label” in their data dictionary csv will populate the DESCRIPTION column in the DD.

⚠️ Issue With nibabies.toml Files🔗

Lasso bug: FE import with duplicated ID impacting the Nibabies file type bids/derivatives/nibabies/sub-<label>/ses-<label>/log/20250125-071817_f14585f4-03b1-48ef-8cdc-edc3eb0be744/nibabies.toml.

⚠️ Missing Labels For 'Levels' On Multi-Select To Check Box Items🔗

There are some unresolved issues with the conversion of multi-select fields to checkboxes for values derived from the Ripple Pregnancy Check up and pushed to the derived Basic Demographics (sed_basic_demographics) directly. The values for the checkbox 'levels' are missing, and only the expected labels (TRUE/FALSE) are available in the Data Dictionary.