diff --git a/easybuild/easyconfigs/c/CSBDeep/CSBDeep-0.7.4-foss-2023a-CUDA-12.1.1.eb b/easybuild/easyconfigs/c/CSBDeep/CSBDeep-0.7.4-foss-2023a-CUDA-12.1.1.eb
new file mode 100644
index 000000000000..1c31a6086b11
--- /dev/null
+++ b/easybuild/easyconfigs/c/CSBDeep/CSBDeep-0.7.4-foss-2023a-CUDA-12.1.1.eb
@@ -0,0 +1,39 @@
+# This easyconfig was created by the BEAR Software team at the University of Birmingham.
+# update: foss/2023a thembl
+easyblock = 'PythonBundle'
+
+name = 'CSBDeep'
+version = '0.7.4'
+versionsuffix = '-CUDA-%(cudaver)s'
+
+homepage = "https://csbdeep.bioimagecomputing.com/"
+description = """CSBDeep is a toolbox for Content-aware Image Restoration (CARE)."""
+
+toolchain = {'name': 'foss', 'version': '2023a'}
+
+dependencies = [
+    ('Python', '3.11.3'),
+    ('CUDA', '12.1.1', '', SYSTEM),
+    ('SciPy-bundle', '2023.07'),
+    ('TensorFlow', '2.15.1', versionsuffix),
+    ('matplotlib', '3.7.2'),
+    ('tqdm', '4.66.1'),
+]
+
+use_pip = True
+sanity_pip_check = True
+
+exts_list = [
+    ('tifffile', '2023.9.26', {
+        'checksums': ['67e355e4595aab397f8405d04afe1b4ae7c6f62a44e22d933fee1a571a48c7ae'],
+    }),
+    (name, version, {
+        'modulename': '%(namelower)s',
+        'source_tmpl': '%(namelower)s-%(version)s.tar.gz',
+        'checksums': ['85d6fc360bb33253ba6f543d75cf0cf123595f0ea4dd1fa76b1e5bc8fc55b901'],
+    }),
+]
+
+sanity_check_commands = ['care_predict']
+
+moduleclass = 'bio'
diff --git a/easybuild/easyconfigs/c/cryoCARE/cryoCARE-0.3.1-foss-2023a-CUDA-12.1.1.eb b/easybuild/easyconfigs/c/cryoCARE/cryoCARE-0.3.1-foss-2023a-CUDA-12.1.1.eb
new file mode 100644
index 000000000000..445545100745
--- /dev/null
+++ b/easybuild/easyconfigs/c/cryoCARE/cryoCARE-0.3.1-foss-2023a-CUDA-12.1.1.eb
@@ -0,0 +1,65 @@
+# Thomas Hoffmann, EMBL Heidelberg, structures-it@embl.de, 2024/05
+easyblock = 'PythonBundle'
+
+name = 'cryoCARE'
+version = '0.3.1'
+versionsuffix = '-CUDA-%(cudaver)s'
+
+homepage = 'https://github.com/juglab/cryoCARE_pip'
+description = """This package is a memory efficient implementation of cryoCARE.
+
+This setup trains a denoising U-Net for tomographic reconstruction according to 
+ the Noise2Noise training paradigm. Therefore the user has to provide two       
+tomograms of the same sample. The simplest way to achieve this is with direct-  
+detector movie-frames.
+
+You can use Warp to generate two reconstructed tomograms based on the even/odd   
+frames. Alternatively, the movie-frames can be split in two halves (e.g. with   
+MotionCor2 -SplitSum 1 or with IMOD alignframes -debug 10000) from which two    
+identical, up to random noise, tomograms can be reconstructed.
+
+These two (even and odd) tomograms can be used as input to this cryoCARE        
+implementation."""
+
+toolchain = {'name': 'foss', 'version': '2023a'}
+
+
+dependencies = [
+    ('Python', '3.11.3'),
+    ('CUDA', '12.1.1', '', SYSTEM),
+    ('SciPy-bundle', '2023.07'),
+    ('TensorFlow', '2.15.1', versionsuffix),
+    ('mrcfile', '1.5.0'),
+    ('tqdm', '4.66.1'),
+    ('matplotlib', '3.7.2'),
+    ('CSBDeep', '0.7.4', versionsuffix),  # cryoCARE 0.3.1 requires < 0.8.0
+]
+
+use_pip = True
+sanity_pip_check = True
+
+exts_list = [
+    (name, version, {
+        'patches': [
+            '%(name)s-0.3.0_relax_requirements.patch',
+            'cryoCARE-0.3.1_fix_np1.20.0deprecations.patch',
+        ],
+        'source_urls': ['https://github.com/juglab/cryoCARE_pip/archive/refs/tags/'],
+        'sources': ['v%(version)s.tar.gz'],
+        'checksums': [
+            {'v0.3.1.tar.gz': 'faf1e06e4a893bf01a6a70b207b75c6ff190310d03fb9b9b56a4be7937935d54'},
+            {'cryoCARE-0.3.0_relax_requirements.patch':
+             'a44814f6e568f5fb618cf789d21a6b5714fbda78b4170ec8e868e50fb0f2a5c0'},
+            {'cryoCARE-0.3.1_fix_np1.20.0deprecations.patch':
+             'a25e3a540016db02d47b82e2104763e376340b80d0d1012ece7bb4a53f8bcdb9'}
+        ],
+    }),
+]
+
+sanity_check_commands = [
+    'cryoCARE_extract_train_data.py --help',
+    'cryoCARE_train.py --help',
+    'cryoCARE_predict.py --help',
+]
+
+moduleclass = 'bio'
diff --git a/easybuild/easyconfigs/c/cryoCARE/cryoCARE-0.3.1_fix_np1.20.0deprecations.patch b/easybuild/easyconfigs/c/cryoCARE/cryoCARE-0.3.1_fix_np1.20.0deprecations.patch
new file mode 100644
index 000000000000..7be1c194054b
--- /dev/null
+++ b/easybuild/easyconfigs/c/cryoCARE/cryoCARE-0.3.1_fix_np1.20.0deprecations.patch
@@ -0,0 +1,18 @@
+# Thomas Hoffmann, EMBL Heidelberg, structures-it@embl.de
+# cryoCARE has requirement numpy ~1.19.2, but we are using newer with SciPy-bundle
+# numpy.bool is deprecated sind v 1.20.0: https://numpy.org/devdocs/release/1.20.0-notes.html#deprecations.
+diff -ru cryoCARE_pip-0.3.1/cryocare/internals/CryoCAREDataModule.py cryoCARE_pip-0.3.1_fix_np1.20.0deprecations/cryocare/internals/CryoCAREDataModule.py
+--- cryoCARE_pip-0.3.1/cryocare/internals/CryoCAREDataModule.py	2023-12-11 10:19:20.000000000 +0100
++++ cryoCARE_pip-0.3.1_fix_np1.20.0deprecations/cryocare/internals/CryoCAREDataModule.py	2024-12-04 10:49:42.527101697 +0100
+@@ -121,9 +121,9 @@
+         
+         # If no mask is specified, just create a one-mask
+         if mask_path is None:
+-            mask = np.ones(even.data.shape).astype(np.bool)
++            mask = np.ones(even.data.shape).astype(np.bool_)
+         else:
+-            mask = mrcfile.read(mask_path).astype(np.bool)
++            mask = mrcfile.read(mask_path).astype(np.bool_)
+ 
+             assert even.data.shape == mask.data.shape, '{} and {} tomogram / mask have different shapes.'.format(even_path,
+                                                                                                                  mask_path)